Skip to main content

tv   Social Media and Combating Terrorism  CSPAN  January 31, 2018 1:19pm-3:53pm EST

1:19 pm
executives testify about what the companies are doing to prevent criminal activity by users and the tech sectors cooperation with law enforcement. just before 4:00 a conference on internet policy and security and live at 7:15 p.m. vice president mike pence speaks at the congressional gop policy retreat in west virginia. c-span's history series landmark cases returns next month with a look at 12 you supreme court cases each week a story ends and experts join us to discuss constitutional issues and personal stories behind the significant supreme court decisions. beginning monday february 25, live at 9 p.m. eastern and help you better understand each case we have a companion guide written by better supreme court journalist tony morrow. landmark cases volume two, the book cost $8.95 plus shipping and handling. to get your copy go to
1:20 pm
c-span.org/landmark cases. thursday morning we're live in montgomery, alabama, for the next stop on the c-span bus 50 capitals to work. alabama senate president pro tempore del march will be our guest on the bus during "washington journal." starting at 9:30 a.m. eastern. >> witnesses from facebook, youtube and twitter talk about what the companies are doing to prevent criminal activity by users and the tech sectors cooperation with law enforcement. the senate commerce and finance committee held a hearing last week. [inaudible conversations]
1:21 pm
[inaudible conversations] [inaudible conversations] [inaudible conversations]
1:22 pm
[inaudible conversations] >> good morning. i want to thank everyone for being a to examine what social media companies are doing to combat terrorism including terrorist propaganda and terrorist recruitment efforts online. the positive contributions of social media platforms are well documented. youtube, facebook and twitter among others help to connect people around the world, give voice to those oppressed by totalitarian regimes and provide a forum for discussions of every political social scientific and cultural strife. these services have thrived online because of the freedom made possible but uniquely american guarantee of free speech and by light touch
1:23 pm
regulatory policy. but as a so often the case enemies of our way of life have started to get revenge of our freedoms to advance hateful causes. violent islamic terrorist groups like isis have been aggressive in seeking to radicalize and recruit over the internet and various social media platforms. the companies that are witnesses represent have a very difficult task. preserving the environment of openness upon which the platforms have thrived while sick and responsibly manage and thwart the actions of those who would use their services for evil. we are here today to explore how they are doing that, what works and what could be improved. instances of islamic terrorist using social media platforms to organize can instigate and aspire are well-documented for example, the killer responsible for the orlando nightclub shooting in which 49 innocent people were murdered and 53 injured was reportedly inspired by digital material that was readily available on social
1:24 pm
media. this issue is not new. over the course of several years youtube hosted hundreds of videos by senior al-qaeda recruiter anwar al-awlaki, although the company promised in 2010 to remove all videos that advocated violence, his call to jihad video in which he advocates for western muslims to count attacks at home remain on-site for years. in fact, the new york times report suggested that his biggest influence afforded terrorist, a boston marathon bombers and the terrorist attacks in san bernardino and orlando. this issue is also international in scope. in response to recent terror attacks in london, which prime minister theresa may has been especially outspoken including social media platforms to do more to combat the kind of radicalization that occurs online. last fall come for exemptions joined by other european leaders and calling upon social media companies could move terrorist content from their sites within one to two hours after it
1:25 pm
appears. as with her today the companies before us are increasingly using technology to speed up their efforts identify and neutralize the spread of terrorist content. in a recent blog post facebook said artificial intelligence now removes 99% of isis and al-qaeda related terror content even before they can be flagged by a member of the community, and sometimes even before they can be seen by any users. youtube is teaming up with jigsaw, the in-house think tank of googles and company of the bit, to test a new method of counter radicalization referred to as the redirect method. seeking to redirect or refocus potential terrorist at an earlier stage in the radicalization process, youtube offers users searching for specific terrorist information additional videos made specifically to deter them from being radicalized. a little over a year ago facebook, youtube, microsoft and twitter committed sharing a database of unique fingerprints of some of the most extreme
1:26 pm
terrorist produced content used for influence or recruitment. by cross sharing this information terrorist content on each of the host platforms will be more readily identified, hopefully resulting result in d more efficient deletion of this victory. a century these companies are claimed the contact individual videos and photos and using automation, can kick them off the platforms before they are even seen. we all have a vested interest in their success, and i believe this to me has a significant role to play in overseeing the effectiveness of their efforts. i do what you think ms. bickert, ms. downs, and mr. monje are being here as represent is for the companies. to mr. watts, i look for to hearing your thoughts about disrupting and defeating terrorism. with that i will now recognize the ranking member senator nelson for any opening statement he would like to make. >> mr. chairman, within a few hours of the pulse nightclub shooting i was there on south orange avenue in orlando.
1:27 pm
and i just want to comment that when a great tragedy occurs such as that, encouraging that the community comes together, like orlando never had. that saying you can say for boston, and so many other places where these tragedies occur. and yet we need to get at the root of the problem, which the chairman has outlined. it's the first time that the commerce committee has had three of the largest social media companies before us. these social media platforms, and those of many other smaller companies, have revolutionized the way that americans communicate, connect, and share information. and by the way, a comment that the chairman made about
1:28 pm
artificial intelligence screening out most of the bad guys stuff. i wish one of you would explain that. that is encouraging, but it's not quite enough, as the chairman has outlined. but at the same time these platforms have created a new and stunningly effective way for nefarious actors to attack and to harm. it's startling that today at terrorist can be radicalized and trained to conduct attacks all through social media. and then a terrorist cell can activate that individual to conduct an attack to the internet, creating an effective terrorist drone, in effect, controlled by social media.
1:29 pm
so thank you, all of our witnesses, for being here in helping explain this. and particularly explain what you are doing to rally to the common defense of our people and our country. because using social media to radicalize and influence users is not limited to extremists. nation nationstates, too, are exploiting social media vulnerabilities to conduct campaigns against this country and to interfere with our democracy, and then on basic values, we have seen the attacks utilizing social media that in up in child trafficking and so forth and so on. now, the russian hackers at
1:30 pm
vladimir putin's direction, attempted to influence, and did influence, 2016 presidential election through all of these things that we've been reading about for over a year, and we also know that putin is likely to do it again .. will seek to influence future elections. and i will be asking mr. watts if you would outline what he sees is happening in this 2018 election. this should be a wake-up call to all of your companies.
1:31 pm
indeed, it should be a wake-up call to all americans. and this should be, regardless of party. this was an attack on the very foundation of american democracy. we welcome the expert that each of you bring to the table today. we welcome mr. watts and his expertise. over many years of how bad actors like russia used the internet and social media to influence so many things. not just elections. and we even know that vladimir putin is reaching down, deep into our government. not just at the top. remember a few weeks ago, part of the net neutrality proceeding, have a million
1:32 pm
comments were traced to russian ip addresses. that is shocking. that is concerning. we should want to know why these comments were filed. what were they trying to do? and all of us should be very concerned about what is going to happen next. in the end, the basic questions that we want to ask is what have we learned? what are we correcting? what is going to happen in the future and how can we get ahead of it before it does? thank you mr. chairman. >> thank you senator nelson. we do appreciate the great panel of witnesses we have in front of us today. thank you all for being here. on my left inherently of the head of -- counterterrorism
1:33 pm
facebook. the head of public policy and public relations youtube, the director of health public policy and philanthropy us and canada on twitter and the senior fellow of foreign policy at the research institute. i will start with one and will move across the table. if you could, confined your oral statements as closely to five minutes as possible. any additional comments will be part of the record but it will give us an optimum amount of time for members to ask questions. thank you so much for all of you for being here and we look forward to hearing from you. >> thank you. chairman, ranking member nelson and distinguished members of the committee. i am monica and the products policy and counterterrorism for facebook. i am also a former federal prosecutor. having spent more than a decade as an assistant us attorney for
1:34 pm
the department of justice. the issues were discussing here today are of the utmost importance. on behalf of facebook i want to thank you and your leadership are seeking more effective ways to combat extremism, crime and other threats to the national security. we share your concerns about terrorists use of the incident. that is why we remove terrorist content as soon as we become aware of it. it is also why we have a dedicated counterterrorism team of people working across our company. this includes experts with former academics, spent their careers studying terror groups. tracking new trends and tactics. it includes former intelligence and law enforcement officials. and prosecutors who worked in the area of counterterrorism. it also includes engineers. you are constantly improving the technology that helps us find and remove terrorist content. in my written testimony i've
1:35 pm
described these efforts in more detail. i also want to note that we pursue this goal with the mindset that is important to maximize free expression while keeping people safe online. we work proactively to keep terrorist content off of facebook. merely through the use of automated systems like matching and textbased machine learning. now, more than 99 percent of isis and al qaeda propaganda that we remove from the service is content that we identify ourselves. before anybody has flagged it for us. once we are aware of a piece of terrorist content, we remove the vast majority of subsequent uploads within one hour. we do not wait for these global benefactors to put content on basement before putting into our detection system. rather, we work with outside experts that track propaganda relief bracelets across the internet and send it to us and
1:36 pm
we proactively put in our system. often, this means we are able to stop the content from ever being uploaded to facebook. however, much of this work cannot be done by machine alone. accurately, removing terrorist content often requires a person. a photo of an isis fighter for instance, that could be shared by somebody who is a supporter of the group, can also be shared by a journalist who is raising awareness for a member of a civil society group is condemning violence and we need people to be able to assess that and tell the difference. we now have more than 7500 reviewers who assess potentially violating content including terrorist content in dozens of different languages. and by the end of 2018, we will more than double the 10,000 people who are already working on safety and security issues
1:37 pm
across our country. some of these people are responsible for responding to law enforcement requests. we appreciate the critical role that law enforcement plays in keeping people safe and we do want to do our part. i was global team responds to requests from law enforcement consistent with applicable laws and policies and this includes responding to emergency requests. where we strive to respond within minutes. we also want to do our part to stop radicalization. and disrupt the recruitment process. that is why we have commissioned multiple research efforts over the past three years. to understand how online speech can most effectively counter violent ideology. we have sponsored efforts to put those into practice. one such example is the peer to peer challenging extreme is a program which we sponsor with
1:38 pm
adventure partners. and through that program, we have had more than 6500 students participate. they have created hundreds of campaigns that have been viewed worldwide more than 200 million times. no one company can combat the terrorist threat alone. we partner with others, including companies, civil society, researchers, and governments around the world. among other things, we work with 11 other companies. including those here. to maintain a shared industry database of unique digital cans of terrorist content so that we can all find and remove it faster. and we have also recently launched a global internet forum where we can work with smaller companies to help them get better. in conclusion, let me reiterate that we share your goal of stopping terrorists from using social media and we will keep getting better at it. i'm here today to listen to your ideas and concerns and to
1:39 pm
continue these constructive dialogues. thank you for the opportunity and i look forward to your questions. >> thank you monika bickert. >> thank you, senator. >> chairman, ranking member and distinguished members of the committee. thank you for the opportunity to testify at today's hearing. and for your leadership on these difficult issues. i serve as the global public policy only for you too. at youtube, we believe the world is a better place when we listen, share and build communities through our stories. our mission at you tube is to give everyone a voice to shut it will purely share over 400 hours of video every minute. with this comes many benefits to society. unparalleled access to art and culture. news and entertainment. educational materials, a remarkable diversity of viewpoints and the freedom to exchange ideas. we value the openness. it has democratized how stories
1:40 pm
and whose stories get told. we are aware however, that the very platforms that have enabled societal benefits like this may also be used by those who wish to promote ticket or extremism. to that end i am pleased on this opportunity alan the approach we are taking on the issues. we have developed rigorous policies and programs to defend against the use of our platform to spread hate or incite violence. you tube has long had policies that strictly prohibit terrorist content. this includes terrorist recruitment, violent extremism, incitement of violence, glorification of violence and videos that teach people how to commit terrorist attacks. we apply these policies to violent extremism of all kinds. whether inciting violence on the basis of race or religion or as part of an organized terrorist group. we use a mix of technology and humans to remove these content
1:41 pm
quickly. users can alert us to content they think may violate our policies through a flag found below every youtube video. we have teams dedicated to this 24/7 in multiple languages and countries around the world. we also work closely with members of our trusted flagler program. ngos that provide highly actionable flags. -- of course, we rely on our technology. which has always been a critical part of our solution. our image matching techniques for example, can prevent dissemination of violent content by touching real uploads of bad content before it becomes public. nonetheless, given the evolving nature of the threat, it is necessary for us to continue enhancing our systems. over the past year in particular, we have taken several steps to build on our efforts. the first is an investment in
1:42 pm
machine learning technologies for the detection and removal of violent extremist videos. we recently deployed classifiers that detected new terrorist content and flag it for review. machine learning is now helping our human reviewers remove nearly five times as many videos as they were before. today, 98 percent of videos we were loath for violent extremism are identified by our algorithms. second, we are focused on improving and expanding our expertise and resources on these issues. we expanded our trusted program to an additional 50 ngos in 2017 including several counterterrorism experts. working with these organizations help us to identify emerging trends and understand how these manifest and evolve. in 2018, we will have 10,000 people across google working to address content that might violate our policies. finally, we are creating
1:43 pm
programs to promote counter speech on our platforms. i will create is for change program supports youtube creators who are tackling issues like extremism and hate by building apathy and acting as role models. the jigsaw group has deployed the redirect method. uses targeted ads and youtube videos to disrupt online radicalization. we also collaborate across the industry. in 2016, we created a sharing database with facebook, microsoft and twitter where is your digital thing offensive terrorist content to stop the spread across platforms. we added seven companies to the coalition in 2017. our shared database now contains over 50,000 video and -- industry collaboration research, knowledge sharing and technology. no single component or party
1:44 pm
can solve this problem in isolation. to get it right, we must all work together. we understand the importance of speed and comprehensiveness in our work. since june, we have removed 160,000 videos and terminated 30,000 channels for violent extremism. we have taken down nearly 70 percent of violent extremist videos within eight hours of upload and nearly half within two hours. we have reviewed over 2 million videos to make sure we are catching and removing all videos that violate these policies. we achieve these results through enhanced enforcement by machines and people and collaborations with outside experts. we are deeply committed to working with law enforcement, government, the tech industry and ngos to protect our services to be exploited by bad actors. look forward to continued collaboration with the committee as it examines these
1:45 pm
issues. thank you for your time. i look forward to your questions. >> thank you juniper downs. >> thank you, chairman and ranking member nelson. distinguish members of the committee and staff.i'm here on behalf of twitter. an open communications platform.it allows more than 330 million users to see what is happening in the world and to share viewpoints from every side. each day, we served 5 million users who have about 3700 employees around the world. twitter has been at the forefront of preventing terrorist explication of the -- as new challenges emerge, we will continue our efforts to ensure terrorist do not have a place on twitter. while also giving voice to those who promote a positive message for the future. twitter has a zero-tolerance policy with terrorist content. this includes not only specific threats of violence, but also
1:46 pm
promoting terrorism , deleting with violent extremist groups and glorifying violence. our job is to enforce this policy globally and at scale to stay one step ahead of terrorists. we have dramatically improved our ability to implement these rules and have suspended more than 1.1 million terrorist accounts since mid-2015. our progress fighting terrorist content is due to the commitment to innovation. while there is no magic algorithm, for identifying terrorist content, leave increasingly improved the effectiveness of our in-house proprietary technology. our technology supplements user reports, human review, and significantly augments our ability to identify and remove content from twitter. at the beginning of 2015, our in-house technology detected roughly one third of terrorist accounts that we pulled down at that time. last year, the number increased dramatically. they identified it more than 90 percent of suspensions for terrorism by our internal
1:47 pm
tools. 75 percent, three quarters of those accounts were suspended before they had a chance to treat -- tweet. of course, like any determined adversary, as we make it part of terrorists to use twitter their behavior evolves. escapement of this we continue to invest in technology to invent new counseling opened to replace those that we suspend. while also developing further, tools that prevent the distribution of propaganda in the aftermath of attacks. because this is a shared challenge, no industry has established the global internet forum to counterterrorism. which is focused on learning and collaboration on technical cooperation and research. twitter see this as substantial opportunity to ensure smaller companies are not soft targets for terrorists. the engagement 60 smaller
1:48 pm
companies over the last several months to share best practices and learning. we plan to grow in that work. removing a tweet does not eliminate the ideology behind it. so we invest highly in alternative narratives. twitter has participated in more than 100 ngo trainings and events around the world. since 2015. we work with respected organizations to empower credible, nongovernmental voices against violent extremism. as part of a continuing effort to make twitter a safe place for open democratic debate, late last year we broadened our rules to prohibit accounts affiliated with violent extremist groups. and to make hateful imagery much harder to find in our platform. we also stepped up enforcement of abuse reported by witnesses and increased transparency about our enforcement decisions. to further educate our users of our terms of service. twitter has also devoted significant resources to combat this information and election interference by foreign state actors. to prepare for the us midterm
1:49 pm
elections, this year, the cross functionalb& test versus plan t verify major party candidates as a hedge against impersonation to maintain open lines of communication with federal and state election officials. to continually improve and apply our technology to adjust networks of malicious automation. and to moderate conversations related to the elections. the company is here today have both shared and unique challenges. while we are competitors in the marketplace, we are close partners in combating the threat of extremism and those who would harm our democratic process. thank you, thank you peer leadership and i look forward to this discussion. >> thank you carlos monje. >> chairman, members of the committee, thank you for having it today. 10 years ago al qaeda in iraq videos on youtube. then al-shabaab played out on twitter shortly after facebook groups and twitter feeds brought islamic states the
1:50 pm
world attention and into the homes of new recruits before they scurried off to other social media platforms like telegram.four years ago most global jihad social media i stumbled into russian influence campaign. the reboot of an old playbook -- social media at its height can offer the platform for discussion across diverse audience and chandler bad actors will always come to these information gateways to pursue objectives. less educated populations around the world dominantly arriving in cyberspace to be a mobile phone to be particularly vulnerable to social media manipulation of terrorists and authoritarians. american focus on islamic state social media equipment for russian modeling in the presidential elections in 2016 overlooks other indicators of damaging activity. american companies have suffered and been a particularly vulnerable to smear campaigns launched by foreign state actors there
1:51 pm
malicious false narratives pushing the stock prices down and decreasing sales with reputation damage. beyond smear campaigns and character assassinations, the committee should take seriously the ability for foreign nations to mobilize violence in the us during ablation i would call -- just a few years ago, appetite is external operations leader in yemen recognize the power of the internet to recruit and mobilize terrorists in america to conduct violence in the as homeland. the islamic state took this to another level with their spokesman. calling on supporters to conduct attacks at home and then further enabling by using a social media patella to guide them remotely. a little over one year ago americans on individual consume a false conspiracy on the internet and social media known as pizza gate and then traveled to washington d.c. to investigate. he rented a falsely implicated restaurant industry as a weapon before being arrested. surely, an adversary theorist
1:52 pm
sees an opportunity in combining these two scenarios. the greatest concern moving forward by a likely foreign intelligence service posing as americans on social media and infiltrating one or -- social media companies will be better positioned to stop this potential scenario from occurring. then us intelligence, security that are applying technical signatures. they raise to implement policy to prevent the last information attack have yet to anticipate the next abuse of the social media platforms by emerging threats. i've offered a range of recommendations for bad actors using social media and previous testimony. i will focus on some issues today. the first and most pressing challenge comes in the debate over social media accounts an anonymity. this allows them to speak out about injustice but over time
1:53 pm
anonymity has empowered hackers and extremists and authoritarians. under the veil of anonymity, and express hate, recruit members, advance divisions in american society. social medias can and should protect anonymity of a z if they choose but they must determine a real person resides behind it. social media is becoming better advanced with tools recently that certify authenticity. however, current level is sub optimal. i encourage twitter to rapidly expand verification to as many users as possible, as quickly as possible. also we are connected to the issue of counter authenticity is the rise in propaganda. the anonymous replication of account that would see the broadcast high-volume misinformation pose a serious risk to public safety and when employed by authoritarians is a direct threat to democracy. lastly, social media companies continue to get beat because they rely too heavily on technologists and technical detection to catch bad actors.
1:54 pm
artificial intelligence, machine learning will greatly assist in cleaning up nefarious activity but for the near future will fail to detect that which is not been seen before. those with intentions and actions of terrorists, criminals must act alongside technologists.i know it is unreasonable to think that every social media company can and should hire threat analysts for every possible emerging threat. the variety of rapid outreach approaches with extra associated media experts, could easily be developed and be collectively sponsored by social media companies. several models from counterterrorism cybersecurity to be adopted by silicon valley in this regard. i've made other recommendations in the past which i can address during the q&a. some social companies have others for the integrity of the plot have a lot to do to
1:55 pm
improve the american consumer will decide whether benefits of using these services outweigh the risk many are working with social media applications because they cannot trust the information being shared or tolerate that user experience. social media companies should move aggressively. not only because it is best for the users in society but because it is good for business as well. thank you for having me. >> thank you clinton watts. >> thank you for the introduction and we have a new member of the committee. senator jon tester of montana. senator tester, has been wanting to get on this committee for quite a while. and we are so happy that he finally was able to be appointed to the committee. he brings a wealth of experience as a senior member of the appropriations committee. to this committee. welcome senator.
1:56 pm
>> mr. chairman, quickly, first of all thank you for the welcome. i look forward to work on this committee although i think i have flashbacks of 2007 right now. thank you. >> i also want to thank you and welcome senator tester for enjoyment committee. it is always nice to have another little senator on the committee. someone who is my neighbor and represents a state like in the mind where there are more cattle than people. and obviously, an area where there is still cell phone and broadband free areas and we are hoping to change that. he can remember times when i can that there were party lines. we have a long way to go and issues that we address on this committee are those impacted daily lives of people in the state. just like they do so many others. senator, welcome. it is good to have you here. thank you. we will start with some questions. i want to direct this to mr.
1:57 pm
carlos monje and monika bickert. google, facebook, twitter and microsoft announced the formation of a hashing coalition to better identify and ultimately remove offending content. the question is, is there any shared standard for what constitutes extremist for terrorist content in your coalition? >> thank you chairman for that question. all companies are possibly working with one another and small companies to address the issues that change and evolve. as new things that we see around the world, we are constantly adapting how we attacked the challenge and we do rely on the advice of the council. not only of our peer companies but academics and ngos. >> there is no standard definition though.
1:58 pm
that you have agreed upon. >> that is right, mr. chairman. i would just end of companies -- we launched the global internet forum in june 2017. but we have been working together for a number of years informally. part of those meetings is discussing what the appropriate standards are. recognizing of course, that these different products work differently. but there are two types of policies i think you most commonly see. first are directed towards the groups having any presence on the platform. for instance, if you're on facebook and both are wrong, you cannot have a page on facebook even if you're talking at the lovely weather. we simply cannot be on that platform. the other types of policy that you see across major companies, banning any praise or support of these groups for their actions. >> thank you. ms. juniper downs, according to extremist project one single bomb making video used to instruct the manchester suicide
1:59 pm
bomber has been uploaded to youtube and removed 11 times. but continues to resurface as recently as this month. how is it possible for that to happen? why aren't your hashing efforts working to keep this video off of your platform permanently? >> thank you, chairman. as i mentioned, we have strict policies against terrorist content. including videos that instruct people on how to connect terrorist attacks. certainly this includes bomb making videos, but as instructing people on how to drive vehicles into crowds and so on. this particular video was caught by our system. we have used and are catching the uploads of the video quickly and removing it as soon as the uploads are detected. >> are your companies, anyone can answer this. are your company is as you start to rule out some of these new counterterrorism programs, how do you have ways of measuring their effectiveness?
2:00 pm
what is the metric or standard? >> chairman, at twitter, we really double down on the technology. on the machine learning to try and identify and remove content as quickly as possible. so our metric is how many accounts are we taking down, have accounts are we identifying proactively and how many are we able to take down before they are able to tweet. ... it has been reported that isis surrogates were using ai fox to carry out recruiting and messaging campaigns as it will
2:01 pm
become more sophisticated and how to prevent and root out the bad people also become more sophisticated in how to get around the threat involves. are you seeing that level of sophistication and if so, what are you doing to mitigate it? >> the use of ai against you by these groups -- does anyone want to take a stab at that? >> we have very aggressive proprietary standard sections systems which would catch massive re- uploads of ai generated videos. our long history in fighting spam across our services are in the technique to get at that figure. >> anyone else? >> i would disagree with you, german that it is a cat and mouse game and were constantly evolving to make the challenge. when we were in the past fan
2:02 pm
accounts would try to come back and brag about the fact that they were being. that became a strong signal for us in which resulted in them being taken down quicker. they stopped doing that. >> my time is expired. >> esther watts, i would like to take my time and inform the committee with your expertise what the russians for that matter anyone else can do to us in this coming election. >> thank you, senator. there's been no response from the us government with regards to russian influence campaigns in social media. therefore, they have stayed on course with their operations enduring nonelection years they can focus on social issues and audio infiltration.
2:03 pm
any organization, entity, social media group that they want to for influence later they begin infiltrating the same content that with that audience and trying to develop their own content within it. beyond the united states in this presidential election i think we should look at all elections worldwide. they realize the playbook works well and it's cost-effective and there's been almost no downside to this point of doing it. you have seen in europe where they continue to see audience bases anywhere they can break up the union so the european union or nato they will continue to see and look at catalonia or scotland or places like they've seen breakup an alliance or create divisions in a democracy there been there. i think general mcmaster last week pointed to mexico as another example where the scene some sort of audience infiltration. the key trigger i always look for his hacking. when they launch widespread
2:04 pm
hacking against a target there making a strategic decision to go after an objective and that's one thing i would tell everyone to look forward on the horizon. the on that if you want to know where the russians are going with their influence you should always look at where they are putting up new state-sponsored outlets to detroit and audience you have to have a piece of content and to launch a campaign so when they add additional language or wire service and let satan use or youtube channel that is an audience they will reach for and i can tell you right now they're looking very heavily into latin america. i think they would like to build a capability more in the middle east moving forward. beyond russia as they focus on social issues to win over audiences during election years and to permit them toward whichever candidate or party they want to support moving forward. the goal is in one party or another but to create divisions inside the united states. that would be the predominant
2:05 pm
before. further polarizing the information landscape i would also note that everyone is adopting this technique. you see it in me and mark, philippines and any low-level, educative operation around the world that is on social media, particularly through mobile application is highly portable to this. they're not built up the ability to assess information or how to avoid being influenced and so they are highly vulnerable to this influence technique. lastly, i would say it's political campaigns in the companies that are going to be hired. if there is not some regulation around as in social media every political campaign whether the us or the world will have to use the dark social media campaign to super pacs or candidates to keep up with their competitors. it will further and not only harm the society within but harm the social media companies and their problems.
2:06 pm
they will actually make the environment so terrible and so polarized as we've seen over the past two years that it will create a nasty sense of democracy and if you want to look at how this affected in russia, russia did this to their own people first before the game across the ocean. it creates widespread apathy in democracies. it dilutes the line between fact and fiction and when it happens you cannot keep the democracy we heard. i think that is what most dangerous about the system is the agnostic party or candidate ultimately it's about breaking our system and turning it against each other. >> so when you see them dive deep into the instrumentality of government such as the example i gave that there were half a million comments on the recent fcc rule and when you see that you read the public press that
2:07 pm
they are in 20 states elections divisions sketch out what are some of the dastardly things that they could do to undermine america. >> the one thing they would try to do is an information tacked on democratic institutions. it's playing out in both of those institutions. the sec if you can't trust the sec we need to get rid of regulatory moderators and you can trust them there trying to mind control you. the other part is the election for the integrity of an election. the second campaign they launched in the run up to 2016 was voter fraud election rig. they didn't really care what candidate wanted but they wanted the american people to think they were both in that account. the hacking campaign against voter databases was to sow doubt that when you see the narrative of voter fraud election rig you might think maybe my candidate didn't really get elected because my vote didn't count.
2:08 pm
it's about destroying democratic institutions in the us government or democratic institutions to govern properly and the system is always rigged in can't trust anyone. that is the focal point of all of those efforts and the russians might run or any authoritarian regime that wants to run a campaign against us government. >> thank you. senator nelson, senator wicker. >> it wasn't so much an attempt to get one candidate elected over the other but it was knowing there was going to be a loser and that relatively half of the population would think their vote didn't count. >> that was their second campaign. they iran for narrative during the first one which were specific to the candidates there in october and it shifted to the integrity of institutions where it was to. they were running to try to get a candidate they wanted up to election day and then beyond election day it was to create mass chaos inside the united
2:09 pm
states. >> thank you for clearing that up. let me move to twitter. if i am and pronouncing it crack, okay, well is that a cajun thing, you tell me you're from new orleans. >> i am from new orleans. i'm a saints fan but my family is from argentina. >> very interesting. let me ask you about aggregate user data. there are analytics companies who purchase aggregate user data from debt and from including twitter, is that correct? >> yes for example, if i am an analytic company and i want to work for the nfl, for example, i would purchase aggregate user
2:10 pm
data from twitter and using keywords develop information that might be helpful to say the national football league. >> it depends on what it is you expand to use it or a lot of times it's most often used for the target advertising. >> so target advertising. okay. let me ask you this. is that same ability to purchase aggregate data available to federal law enforcement and is this available to federal antiterrorism agencies? >> it depends on the purpose of user data. we have rules about any entity regardless government or not anywhere in the world can user data. >> okay. what are those rules with regard to terrorism?
2:11 pm
>> i'd have to get back to you on the exact language. >> okay, because this is pretty important. if an analytics company wants to purchase data from twitter you're willing to sell that to them customer what i want to know is if that company will supply information to agencies that are seeking information about terrorist activities and that activity is part of that aggregate user data will you sell that data because frankly i'm informed that you will not do so. >> let me tell you a little about what we do with our data service. we, on our site, on the tender side of the equation are very data focused user data to inform the machines that we fight the terrorists we work on a daily
2:12 pm
basis want. >> okay, that's within within the twitter organization. >> yes, you work on a daily basis with law-enforcement particulate with the fbi and respond to any request they have as long as they give us the upper legal process and we are on a first name basis with our counterparts at the fbi. >> what was that proper legal process be? >> it depends on what they're looking for. it could be a warrant whether they're looking for private or non- private information whether they're looking for direct messages. >> has put her told these data inlet companies that the purchases of this data cannot be used for counterterrorism purposes. >> it cannot be used for persistent surveillance of individuals. >> they can be used to target advertising and products in sales but they cannot be used to
2:13 pm
help our antiterrorism agencies? is that correct? >> we do help our antiterrorist activities particularly desperate. >> but if. >> i'm sorry, sir. >> go ahead. >> we have a very fast system that is an input in anytime they have information to us we turn it around as quickly as we can within hours. we do not -- >> you protect the privacy of your users even if a federal agency wants to use wants to surveilled that public information for antiterrorism purposes? >> if an agency comes to us with the right process and its according to federal law. >> but that is not what i'm talking about. >> well.
2:14 pm
>> i'm talking about an independent data analytics company. you will sell that data but you'll tell them you can't use it for terrorism purposes. >> will not allow any company whether cars or serial or anything, the nfl, to persistently fear out were someone is at a given time but we do have data alerts for law enforcement that the fbi use -- >> is that the policy of facebook? >> thank you, senator. we don't sell user data outside the context of allowing people to target audiences in their advertisements for that is the capability that is equally available to law-enforcement as it would be to anybody else. law enforcement can provide us, if they want to find out
2:15 pm
specifics about an individual user, they can provide us with legal process and will respond. >> what is the privacy concern that supersedes the need to surveilled terrorist organizations that participate in this book? >> senator, we absolutely respond to valid long request. if this is a part of an investigation and they give us the process we do respond. >> thank you, mr. chairman. >> thank you senator wicker. senator. >> thank you. i came in and three senators told me you were a state. is that correct? [laughter] i would like to note my scarf and won the game, who won the game? >> it was an excellent game in a spectacular and for the vikings. >> okay, to give her much. now, let's get to serious matters. we've had a hearing focused on the election of the internet and
2:16 pm
i noticed that i've noticed questions about terrorism as well. there are many ways we can undermine our country and undermine our democracy and it's obvious with violent attacks and another is the american are able to make their own decisions about who they were voting for because they get false information and that is why senator warner, mccain and i have introduced the act about i would start with you, mr. watts, right now there is exposure rules and radio, tv, print, they all have to keep on file as a national political important legislative importance as well as candidate as so that opposing can see these acts as well as disclaimer requirements. do you think those apply to social media ads, paid ads as well? >> absolutely.
2:17 pm
it does not happen both for society and social media perspective the conversation will continue to get more polarized and negative. people won't be able to trust information on the platform regardless so i think it is essential that the ad regulation extend social media because that's where it's going in the future. >> we are $1.4 billion in the election and their protections it will go to three or 4 billion and things like that and there are literally no rules. we appreciate a number of companies included stepped up to start putting in place some of their own guidelines in changes but i do believe this will not work unless we have guidelines like we have for medias. do you agree with that? >> i do. don't it could have a devastating impact. all political campaigns try to do social media manipulation and maybe not entirely augmented. >> thank you. my state has had its share of
2:18 pm
better greeting as especially related some from isis and in past years had dozens of prosecutions under us attorney's office where ones have people have been recruited to go overseas to fight on behalf of of terrorist groups. what kind of recruiting activity are you able to detect under platform and what can you tell us about the trends and how are they changing the strategies? i remember the fbi showing me the ads targeted at minnesota where airplane tickets from minnesota to somalia and tell me what you're doing now what you see in terms of recruiting and what you can do about it -- anyone know one -- should i call people? >> i'm sure they don't want to answer is much as i do so i'll go first. we should notice the social media companies were a forerunner but there also the time stores of the social media
2:19 pm
era meaning there's a large platform for the greatest capability for terrorist activity but in the future if i were a terrorist or in extreme his group trying to mobilize i would go to the smaller social media applications that have the greatest encryption, the largest dissemination capability and i would focus their and then moved to other social media platforms because there would be less ability for them to determine deter my activity. with that in terms of the extremists i thank you need to look at what of the social media applications being used by language, language is the key for doing agreement and where the populations in each of your state and the refugee population and immigrant population how does that play out in who are the interviews he was overseas. one last question. the 2016 election they tried to impose electric and part of it was by searching algorithms to promote misinformation. in the current new era
2:20 pm
information is consumed rapidly in algorithms by significant role in deciding what content consumers see. what are twitter and facebook doing to help ensure the information appearing in search results and on consumer seeds is free from that expectation? >> thank you for that question and we do write a bit to protect our search in particular more than 95% of our users as a default setting essay searches as part of what they experience on twitter. we identify a malicious automation which is a lot of the ways that this kind of information has promulgated on the internet is severely downrange his heart. >> okay. >> thank you, center. we are increasingly finding new ways to disrupt bosnia's and
2:21 pm
help people connect with authenticators. we know that is what they want to do and we also invest in efforts to help people distinguish between the two which include basic education and public outreach. as far as disrupting the positives, often times because we have a requirement that people have to use facebook with their authentic names can identify inauthentic accounts and were getting much better at that we can remove those accounts and the policies goes way. the majority of the actors we see trying to spread disinformation are financially motivated so that goes a long way. we are also working with our community, factors and make that content less visible and put it in context so now if you come to facebook and you see a story in
2:22 pm
your new speed that is an article that has been flagged as potentially false community we will show you some related articles underneath it so that you have a sense of where this story sits in the broader spectrum of news. we are working with the responsible publishers to make sure that they know how to most actively use social media and were working on user education. >> senator moran. >> mr. chairman, thank you. thank you for conducting this hearing. it's on the most interesting and potentially valuable hearings we will have had and what a great development it would be to produce the military necessity in the loss of life that comes from military action in fighting terror to keep it from occurring in the first place. thank you for being here and thank you for your testimony. let me ask me ask this.
2:23 pm
some of you covered in your testimony collaborative efforts among multiple businesses and groups that involve shared industry database which eventually led to the formulation of the global intermittent form for counterterrorism. i want to know more about that collaboration and part of the reason for that question is that my guess is that as the larger social media companies become more innovative and effective in what you're attempting to accomplish preventing terrorism it would seem to me that other smaller platforms may become the platform of choice in this space. so if you are successful in your efforts what prevents us from moving terrorist from a different platform and therefore those smaller platforms engagement in what you're doin - directed at anyone or if maybe
2:24 pm
no one does -- >> thank you, senator. that is exactly what we were thinking as a large company. we need to make sure that this movement was industrywide. with that in mind we have reached out to a number of small companies, several years ago and we reach out to eating companies initially in all 18 said yes, they want to meet and talk about best practices to counterterrorism and we then met for more than a year before we ultimately launched the global internet forum. through that form was launched in june we since had five international working group sessions with 68 smaller companies based around the world and this is an opportunity for us to share learning from the larger companies. >> let me take that a step further. one of the smaller problems doing? they're doing something similar
2:25 pm
to what you're telling us that your companies are doing today? >> yes, senator. often they are learning from what we are experiencing as the larger platform in terms of the conduct we see from bad actors in the policies we put in place and how we think about using technology and people to combat. >> anyone else? i would add. >> i would add that we are extremely successful in taking terrorist contact on twitter and is a tremendous success for twitter but doesn't illuminate the terrorist and then moving to other problems like telegrams doesn't help everybody and twitter is a smaller company among the giants and we often because we've had to be creative and innovative ways of technology can help you the bridge to smaller companies and tell them you can make
2:26 pm
significant process you just have to invest in technology. >> what are you seeing terrorist change their behavior as a result of what you were doing? >> there is open in some of their forms are now they are trying to find a platform where they can go in a secure fashion communicate and push their propaganda around the world so they are actively seeking new platforms in your question is a great one which is how do we help the small companies that are developing new social media applications don't have the capabilities in terms of security toward the sauce. i don't think there's a good answer for that question but they are seeking a new home and they just have not done it yet. >> esther what, is the response by terrorist organizations to seek a new home or are they finding new ways to hinder your efforts or both? >> both. they're looking for a place to communicate and organize they have to be up to push propaganda
2:27 pm
globally to recruit and gain resources. they need some way to do that. they will continue to try to exploit these small applications but it is tough for them on the small applications because globally there is not as many people on it. it's a better problem to have the one we've had in the past but it really begs the question that ultimately one of the social media platforms is popular overseas will gain traction with them either due to its encryption capabilities or how they connect with the audience or how they load and share videos. i think this is important across all extremist groups. if you look at some of the problems that are out there read it, these anonymous platforms they could be great tools in it be great to see them integrated to have more capabilities. >> think much. >> senator. >> thank you for all being here. facebook had $10.3 billion in revenue last quarter and about
2:28 pm
3000 employees, if not mistaken for twitter $590 million in revenue last quarter, 3700 employees so my question facebook and twitter is in dollars as a percentage of revenue however you want to cut elated and in terms of employee accounts both part-time and full-time how many people me dollars are you devoting to this problem? >> thank you senator. this has been a significant area of investment for us not just now but over the past years. i do want to point to a recent announcement from our ceo mark zuckerberg after we released earnings last quarter where he specifically pointed to the fact that as we invest in security we should expect to see these books cost go up. that is reflected in the fact that we now have more than 7500 people who are working to review terror content and other potential violations.
2:29 pm
we have 180 people who are focused specifically on countering terrorism so these are people like the former academics like brian fishman, formally with though counterterrorism research center desk. >> so 180, 750 -- 7500 our content repairs. in the safety and security more generally we have 10000 people currently and we are looking to be a 20000 by the end of the year. >> and you? >> it is pure than that. i can tell you our entire engineering product is at various stages are all working on this. we are a small team and we have to be. have to be able to shift as a challenges. the numbers that are important also to look at our 2 million users, 400 hours of video every minute and for us 350,001 every
2:30 pm
minute and this -- in order to make progress on this issue you do need to had humans and we had former law enforcement and we have people contractions, consultants -- >> i want to give you an opportunity to set the record straight about fake accounts. i've been reading a lot about this and saw anywhere from 915% stake and i saw another usc that said there's 48 million users and what is the number, the fake accounts you have? >> we believe that fewer than 5% of the council under our fake accounts. >> know if you have zeroed in on let's stipulate that it's 5% almost 300 million and if you know they are fake what is the issue here? >> we keep going back and they tried different methods to get back on the radar screen and so we are, as a matter of course,
2:31 pm
consistently fighting religious automation and we are now challenging million malicious automated accounts a week and that means we are essentially sending them a note saying you're acting weird can you verify you are a human being. that's double to where we were last year. >> i know the hearing about terrorism primarily but it is worth asking what we're doing about active measures you know, there was public reporting that the way marketing went from 27,247,000 followers and those appear to be located in russia and we had to take anything where clearly there was an active measure to try to sow discord. in other words you got platforms
2:32 pm
out there that are taking both sides of the argument is when we get into a conversation about active measures against our country i don't think we should think of it as active measures against democrats and i don't think we should assume that it is just russian active measures. you have to think of this as undermining democracy itself and undermining our ability for our first amendment rights exercise in any way that is meaningful. the question for you is relatively recent and it doesn't seem -- you can give us the measure of your activities, 4 million accounts are being challenged and 500,000 accounts have been taken down but based on results you are not for you need to be for us to be reassured that you are securing our democracy and the degree and extent to elected officials and people who vote and our adversaries are participating in your platform how can we know that you will get this right and for the midterms? >> thank you for that question. that's the question we ask sell
2:33 pm
70. we think were better prepared for the selection that we have ever been a number continuingly between our tools and we will get better and will report to the american people results of our efforts. >> thank you. >> senator young, is not here. senator markey. >> last month the sec gutted the net neutrality rules that protected the internet as we know it and as a result the next facebook in the next youtube and the next printer will start her to get off the ground. i strongly opposed that fcc decision which is why i plan to introduce a congressional review act resolution of disapproval which will undo the sec's recent actions and restore the 15 open internet border. my resolution enjoys the support of democrats susan collins, senator from maine she will vote for it. my question to each company here is simple. you support my cra resolution
2:34 pm
will put net neutrality back on the books? >> yes sir. thank you for your leadership on this issue. >> he would support it, thank you. >> we support strong and forceful net neutrality protections. we supported the 2015 roles and we will support any effort to put those rules back in place. >> thank you. >> thank you, senator. same answer. we will support cra and we also support will work with anybody who is interested in working to find a way to put those rules back in place. >> thank you. thank you to be. next question, bad actors can and do use the internet and social media to apply and acquire weapons including firearms and in 2016 i wrote a letter to facebook and instagram asking why gun sales continue to take place on their sites even after announcement of the proposed policy changes being
2:35 pm
eliminated and i was pleased when both facebook and instagram announced would prohibit individual users from buying and selling firearms. yet recently media reports indicated that users are still able to gain access to deadly weapons on social media. assessment the chicago police department arrested the people the case involving the sale of illegal guns on facebook groups. it appears the gun sales on your platform may have loomed into private groups. how is facebook working to stop the sale of firearms in the corner of your platform, notably the police department in chicago said it did not receive corporation from facebook during its ten month investigation. lump is beneficial reported that facebook hampered their investigation by shutting down their officers were using to
2:36 pm
treat the group in question. >> thank you, senator. it is certainly an issue we take seriously and is a former federal prosecutor based in chicago relation to law enforcement authorities is very important to us and we have cooperated with law-enforcement and will continue to do so in that case. we do not allow firearms sales and enforcement has presented challenges for us and to get better one of the things we're doing is working on our technology. anybody in the community can report gun sales to us and we will take action and that is important and it does happen even in private groups but we know we need to do more. that is why we are now using things like image protection technology to help us recognize when the sales might take place. >> okay. so since instagram can turn into a monster gun for someone who
2:37 pm
plans on using it for the various purposes the answer you would give to the chicago police department when it said it did not receive corporation from facebook during its ten month investigation is that you did cooperate or that you have now established a policy of cooperation with chicago police department and every police department across the country? >> i believe they clarified their statement "after words". we have been cooperative with them from the beginning and i'm happy and i would be happy to follow up "after words" with you on that. >> that would be helpful. in terms of the private facebook group that this type of activity has migrated to, you are saying as well that you are working to shut that down as well. >> that's right, senator. this is an area where we
2:38 pm
recognize enforcement can be challenging and we have to be proactive in looking for solutions so trying to make it easy for people to report but also going further to lucrative content. >> and that is why this hearing is important. we thank you because the internet can be used to spread hate but also to spread weapons of those who are the haters and do enormous harm. in all of our communities across country. thank you. thank you for your testimony. >> thank you and i would ask the three of you that you would support legislation that would put in place rules for open internet as well. with that be true? >> twitter has long been a supporter of net neutrality and we hope congress can develop good rules. >> same answer.
2:39 pm
>> same answer. >> very good. next up is senator baldwin. >> thank you for this important hearing. much of the conversation today has been this on addressing foreign terrorist organizations, use of your platform, as tools to recruit and radicalize individuals both here and abroad. i would like to turn to how you are addressing the use of social media to further domestic extremism whether it's the particular attack on counter protesters in charlottesville the summer or the 2012 shooting at a temple in my home state of wisconsin. we have seen numerous individuals describing to racist ideologies turning to violence.
2:40 pm
beyond that there is disturbing increase in hate crimes in this country as documented by fbi limited collection of data from state and local law enforcement. as with other forms of extremism social media is undoubtedly playing a role in spreading these ideologies in challenging these individuals into violent action. how do your companies working to address the role of social media in furthering domestic extremism particularly white nationalists and supremacist and i like to start with you. >> thank you, senator. i want to be clear that our policy prohibit any group that is either a violent organization and that is regardless of ideology.
2:41 pm
if it is a domestic terror organization or a foreign terrorist organization or no matter the ideological underpinnings they are not allowed on facebook. we also prohibit hate organizations. these are groups that are propagating hate based on a protected characteristic like race, religion, gender, gender identity and support. the same consequences under our policies are applied. they're not allowed to be on a platform and cannot support th them. >> thank you, senator. our violent extremism policy applies to violent extremism in all its forms including white supremacy and other forms of hatred and we apply our policy against violence and violent ideology consistently across violent extremism and all its manifestations. >> very similar answer, as well. we don't allow violent extremist groups. we don't like glorification of violence and it's also when
2:42 pm
charlottesville was a hard day for a lot of folks and you saw online the small minority of folks who were saying terrible things but the vast majority of folks are coming out to reject it. >> were going to turn to different topics. i'm concerned by president trump and secretary of state tillerson's reluctance to support staff. and the state department's global engagement center. it is tasked with coordinating us efforts to counter extremism propaganda and recruitment as well as russian active measures like this information. i like to hear from each of the companies about their experiences working with the center and how cooperative efforts could be improved and mr. watt, what are the national security impacts of this administration's failure to prioritize the center especially in the context of russia -- why
2:43 pm
don't we again go down the line. >> thank you, senator. we are committed to working with governments around the world in promoting and finding counter speech solution and we have worked the global engagement center and others in the us government. we have found that collaboration to be effective and often what we find is that government can be effective as a convening power and bringing together civil society stakeholders and industry and researchers to get together and share their knowledge. that is something we hope to continue in the future. >> our efforts to combat terrorism on our product obviously start making sure we are removing the most egregious content but an equally important part of the strategy is to do
2:44 pm
the hearts and minds working to address the issues at their roots. we meet regularly with ngos and government actors including the state department and global engagement center to talk about the importance of investing in the work. similar answer, as well. government does have an important role in combating this issue and not only investing in counter speech but those groups that are authentic places in communities. >> i'm absolutely baffled as to why the global engagement center that they receive that mission when i understood in 2016 for the election and senator portman, if i recall, was a leader of that and i communicated with her staff on the russian issue at a bare minimum the us government needs to have a real-time understanding of what russia is doing in social media and the hamilton 60 platform i have tried to provide to the us government directly to multiple agencies and i have previous government since 2014 with the different content and i sit here today and have no answer for you. i don't understand why we
2:45 pm
wouldn't at a minimum, regardless of the outcome of the election of 2016, want to provide intelligent agencies are law enforcement agencies and the department of defense with just an understanding and we don't even have to counter, just an understanding of what russian active measures are doing around the world. there is no excuse for it. i can't understand it. >> thank you very much, mr. chairman. i appreciate you and ranking member pursuing this very important topic. terrorism and social media is a challenging and i think passing subject and i recognize that technology companies cannot solve this alone but they must do more and i think that has been highlighted by the questioning you've seen here today. i'm focusing in my first question similar to senator baldwin i'm concerned about the explosion of white supremacist
2:46 pm
online. in december after years of posting fantasies about school shootings and hate filled racist rants over many interim net platforms and many other identifies a young man took a gun to a local high school and new mexico and killed two students were taking his own life. listening to you i am wondering what can be done in this kind of situation and in the testimony it was highlighted that the effort that facebook is taking to counteract isis and other foreign terrorists but can you speak to the efforts facebook is taking to fight one of the most one of the biggest threats, domestic terrorist like white supremacist. in the situation where you have an individual under various identities taking positions and
2:47 pm
indicating right on the edge of violence what can be done in this kind of circumstance. having run into situations like this for? >> thank you, senator. it is certainly an important issue and we stand against violence in all its forms and we don't allow any violent organization regardless of ideology and if we become aware of the threat of violence credible threats of imminent harm to someone we proactively reach out to law enforcement authority and that is something we have done in cases where we have seen a threat like a shooter or whatever the ideology is it doesn't matter. we will proactively provide that want. >> mr. watts, do you think more could be done here based on the answers you here?
2:48 pm
>> in terms of domestic extremism eyesight with the social media companies in the sense that it's simple to understand where to fall because there's not good leadership from the us government about what a domestic extremism group is. we have a luxury of what you think we could do more there and if we delineate more appropriately as a federal government we can enable the social media companies to effectively draw the line and i don't like the social media companies having to decide what is free speech versus violence meets or extremist versus norm. put them in a terrible position and i don't think it's good for business. at the same point, how do you do that short of a violent threat or imminent threat and to do that we would have to have the equivalent of an fto or foreign terrorist organization program in the domestic content and i'm not sure how we get there. >> and that may be something we should consider is how to urge the government to be more specific here and outline areas where we can work with industry in order to move that along.
2:49 pm
>> even from there are two different playbooks and without the formalization it's hard for any corporation evenly and legitimately to force any regulation on a user or group. >> in your testimony you outlined twitter rules against terrorism that expressly include users and i am quoting here, cannot make specific threats of violence or wish for serious fiscal harm, death or disease of an individual or group of people". i'm curious then what twitter's position on one of the presidents of videotapes were he was plotting slamming a person with a cnn logo superimposed on their face, video appears
2:50 pm
promote serious physical harm to cnn reporters in the context of an alarming increase in violence against reporters in the us. >> thank you for that question. note for users is above the twitter rules and action accounts on any given moment were looking at whether they are trying to do satire or do humor even if it's not successful humor. we also recognize that what leaders do have special voice and it is in the public interest for their constituents to hear from. >> well, i don't think this was humor and i don't think the result and i think if you look at what cnn reporters have said since this there is more violence towards them is more animosity towards them and i think we need to look at it in a context and i would encourage all the companies at the stable to take threats to journalists very seriously. i'm extremely concerned with any
2:51 pm
threats of violence based reporting that the president finds disagreeable our president calling us media outlets the enemy of the people and i think it is up to all of us to safeguard the first amendment. thank you very much, mr. chairman. >> thank you, mr. chairman and ranking member they keep having this testimony. i think to start with you, mr. watts. can you tell me why transparency between and behind who is paying for political ads and issue -based ads is important? >> yes, i think the number one issue with that is public safety. we saw with russian active measures one of the primary things they sought to do was mobilize population regardless of the election and mobilize people to protest or counter protest which can lead to
2:52 pm
violent confrontations. at the same point those advertisements when annotated and noted based on campaign laws give legitimacy to those advertisers the public knows what a real political stance first of all formative little truth or narrative and they have to stand by their actions. that's important for the public to restore trust and faith in the democratic process. >> yes and it impacts credit impact election results? >> it makes it more difficult for a foreign adversary or even a manipulative social manipulator with resources and an extra trying to do character assassination or tear down social movements. >> okay. and this goes to anyone in the industry who wants to answer this. can you tell me why you don't tell us who is paying for the ads, whether political or [inaudible]? wants to answer that. >> thank you, senator.
2:53 pm
twitter is very proud that we of last year announced industry-leading transparency practices for political advertising. >> so do you tell people speak for the arts? >> for electioneering ads, yes sir. >> issue -based ads are a harder nut to crack. it's harder to determine and we are working with our colleagues and with our peer companies to try to figure out what the right way to address those issues are. >> what about the other two can you talk political ads first issue -based ads and paste them. >> were working to put more transparency into the election -based advertising system and taking four steps in advance of the midterm elections. the first is verification. he will require advertisers to identify who they are and where they are from for purchasing
2:54 pm
advertisements. we also will in and disclosures for reno users of who is running election -based ad we will release a transparency report on election advertising purchased through google and police accretive library for the public for all of that for their made public. >> will that release was paid for the ads? i believe it will, yes, sir senator. >> thank you, senator. our answer is substantially similar to my peer companies on the issue of federal election related ads. like was said a political ad or broadly are more complicated but certainly an area where we think increased transparency is important. >> political ads are more completed than issue ads or the other way around. >> center, issue -based content -based ads are hard to find and we are interested in how we can increase transparency and we look forward to talking with
2:55 pm
yourself and other policymakers about it. >> i would say this as an editorial comment. i would agree with the senator. i don't think this is democrat, republican issue but a democracy issue. you guys are smart case. just about everybody that i read writings of tell me that it's not that difficult and they are smart people, too. i would hope that you guys would put pen to paper if that is what you do these days and figure out how you can let people know who is paying for ads. issue -based ads, by the way, are just as urgent as political -based ads because they fall into the political category. i would say that is important. everyone of you said you did not like the fcc decision on the neutrality that came out a month or two ago. during that debate we learned that there were lots of drop,
2:56 pm
the disorder the whole public comment. how will that be stopped the next time we have a public comment period on a rule that's written by an agency? anyone want to answer that? i'm out of time so make it -- i will tell you what, i will not occupy the time of the committee but give me an answer to that in writing when you go back to your folks. this is a really important issue when i just want to say this is a really important issue, from a terrorist point in all the questions were asked before but our democracy is at risk and we got to figure out how to get this done and get it done right and get it done very quickly or we may not have a democracy to have you guys up to hear you out. >> center young. >> thank you, mr. chairman for holding this hearing. youtube went from having 40% of its post takedowns last year being identified by algorithms,
2:57 pm
ai in machine learning to have a 98% today. twitter went from roughly 33% detection of tears accounts to more than 90% of the deductions today again attributive to algorithms, ai or machine learning. facebook has stated that nearly 99% of isis content is detected and removed for anyone even report. what is -- what is responsible for the recent increase in the use of ai in machine learning for this purpose of taking down posts? is it mainly primarily because the new commitment to take down posts by your company's or is it simply that technology is finally a place to be effective or some, nation thereof?
2:58 pm
>> thank you, senator. it is deftly a timely question. these innovations have been happening over the years and we have seen a lot of improvement particularly over the past one-two years at the book and a lot of these efforts have been in place since i joined the company six years ago such as still image hashing but it has gotten better. in fall 2016 was when we finally found video hashing to be sufficiently reliable where we could use it to detect these terror propaganda videos in for some of them like a beheading video that we know by later policy regardless of how it is shared we can accurately identify this upload. we've been trying to do it for a while and not been able to do. another area where we have gotten better is [inaudible].
2:59 pm
we take down bad accounts and they try to take back in that something that for a variety of reasons has been important to the company for years but an area where we made significant progress in the past-one-two years. the final advance i will point to returning to my colleagues is the area of natural language understanding. this is hard and we have many different languages that we support on facebook and when you train these models they have to be trained on suspicious data. this process takes a long time but we are making progress and were now using it in areas where we can import. >> thank you. >> we have always used the next of technology in humans to enforce our policy and as technology gets better we see it doing more of the heavy lifting and detecting the content that violates our policy and needs to be removed. these are reinforcing where humans make judgments about what content by later policy that
3:00 pm
seeps back into the training set of data to teach the class fires an algorithm what they are looking for. the more content we review over time the better and better they get and the more they are able to affect the content that needs to be removed. >> before i turn to, my apologies, i stepped into the room. i would note you referenced human judgment and how that feeds into an algorithm to help make more informed decisions moving forward. ... and is there transparency or will there be transparency about that decision-making process,
3:01 pm
but again that's for another day since i have 47 seconds left. mr. monje. >> just very briefly we approach similar to our peer companies and costly country get ways we use our technology and feeding it so i can tackle, it can tackle increasingly difficult and more nuanced challengers. >> okay. i just note in the remaining time here, i really enjoyed visiting with yasmin green, director of r&d at googles jigsaw group. i'll say that alphabet is doing some really great work there. i look forward to working with all of you to improve how we remove this horrible content from the internet and keep americans more safe and secure. >> thank you, senator young. next up is senator blumenthal.
3:02 pm
>> thanks, mr. chairman. thank you for holding this hearing, and thank you to this really all-star panel for being here today. mr. watts, i find your testimony absolutely chilling. the internet is a potential monster when it comes to extremists and terrorism, and it requires the kind of inventive and robust investment attitude that, in fact, created the internet. i've been reading a book called the innovators by walter isaacson, and it is an inspiring account of how we came to have the internet and social media, involving heroes whose names have been largely lost to history, including some nobel prize winners. but the point that he makes that
3:03 pm
a thing is so relevant to this discussion is that the internet itself is the result of a partnership between private industry and inventors, government, and academia. and those partners are as necessary for this effort in combating terrorism and extremism as it was in inventing the platforms of themselves. i want to join in thanking you for your commitment to net neutrality. i also want to thank you for the commitments that your companies have made with varying degrees of enthusiasm to our anti-sex trafficking efforts, most especially sesta. i want to highlight the
3:04 pm
importance of the nationalist hate groups and extremist groups that have come to pose a very dire extremist threat. you received a letter signed by 19 19 civil rights groups including muslim advocates on october 30, 2017. it was cosigned to facebook, but it is equally applicable to all of your companies. i ask that if he made a part of the record, and i see the chairman has stepped off, but i'm assuming there will be no objection and it would be made part of the record. [laughing] and i'm not willing to yield a part of my time to address that objection. the southern poverty law center has warned it social media has been instrumental to the growth of the alt-right movement
3:05 pm
allowing legions of anonymous twitter users to use the hashtag alt-right to push far right extremism on youtube, for example, it's easy to find, for any anti-semitic content. all of these forms of extremism often white supremacist extremism have been allowed to flourish, and the pose a real and present danger. in the time that i've left, i want to ask about a letter that i wrote to facebook, google, and twitter calling on these companies to individually inform all users who were exposed to false, misleading, and inflammatory posts generated by russian agents. i assume that none of you have any doubt that the russians meddled in our 2016 election and attack our democracy. any question?
3:06 pm
none. and that the investigation of those efforts is not a hope for a witchhunt, that this danger is continuing, and as mr. watts has dramatically and powerfully outlined, and that they will continue to do it unless they are made to pay a price, and those who polluted and cooperated with them are made to pay a price -- colluded. i want to thank facebook for a substantive response in terms of its commitment to providing consumers with an online tool to inform users if they have interacted with russian sponsored pages or accounts. i'm hopeful that facebook will do even more with more robust steps to further increase transparency in the future, but i am very, very grateful for your beginning. i just want to be blunt. i'm disappointed i googles written response. essentially blew off my concerns
3:07 pm
by saying the nature the platform made it difficult to know who has viewed its content. i look forward to responses from twitter and others. if you want to respond now, i would be eager to hear what your response is to the letter that i wrote. >> thank you, senator. and we've briefed your staff on her plans and will be rolling out the fulsome response shortly. >> and what will that response be? >> we will be working to identify and informed individually the users who may have been exposed to the ira accounts during the election. >> thank you. i think it's so tremendously important that we have all hands on deck in dealing with this threat, not only the companies who are represented here, but again, as mr. watts said so
3:08 pm
well, some of the smaller actors, some of the newer ones, and there will be others coming that provide in the fact platforms for hate, extremism, terrorism, division, chaos. in some ways they are the biggest threat to our democracy today. those groups that want to foster hate and, of course, the russians will continue. they have an asymmetric advantage here. it's an absolutely wondrous investment for vladimir putin. he gets more return on the dollar than any other investment he can make in sowing chaos and discord in our democracies, and we must be as inventive as the innovators were, the inventors of the internet, in combating this threat to our democracy. thank you, mr. chairman. >> senator cortez masto. >> thank you.
3:09 pm
thank you. welcome and thank you for this conversation. i want to start with, unfortunately, a horrific tragedy that occurred in my hometown october first, 2017. las vegas experienced the worst tragedy that we've ever seen and it is the worst mass shooting in american history. as we were dealing with the horrific tragedy of the situation and trying to gain information, particularly along for the purposes, unfortunately a lot of information -- misinformation was pink spread after that tragic on some of your platforms and on the internet, and particularly misinformation about the shooter was highlighted on both google and facebook. that's incredibly unhelpful for law enforcement, particularly as we move through an unfolding potentially dangerous situation. i know both facebook and google cited the need to make algorithm improvements to fight the spread of fake news during a crisis. what do you see as your
3:10 pm
companies roles in fighting fake news, especially during a crisis such as ms shooting or terrorist attack? what specific and verifiable metrics can you provide is to ensure our trust in these remedies? and i will start with ms. bickert. >> thank you, senator. what happen in las vegas was horrific, and there were false news stories that we saw that we did address, but not fast enough here and it's an area where we're trying to get faster. we exchanged the way that our crisis center operates so that we can make sure that that type of false news story does not appear in the headlines that people are seeing. the crisis center can be incredibly useful during times like this. in las vegas we saw people using not only our safety check, which allows people to say that they're safe, but also coordinating health, offers of
3:11 pm
housing and assistance to people throughout the city. so we want to make sure that's working effectively. things we're doing, removing the bad accounts that are propagating this false news, making algorithmic changes to make news that is likely to be false less visible on the site, providing related articles. when people see a new story that has been flagged as something that might be false so they can see the broad spectrum of information across the internet, and in working with responsible publishers to make sure that they know to use our tools to get their stories out there. >> thank you. ms. downs? >> thank you, senator. my heart goes out to the city of las vegas and all the victims of the senseless tragedy. we take misinformation on our platform very seriously and we made a lot of efforts in our products from improvements to our ranking algorithms to highlight authoritative sources into demote low-quality or less reliable sources, particularly when users are seeking news content.
3:12 pm
we also a strict policies in place against the monetization of new sites that are misrepresented themselves in order to remove the financial incentive to create and distribute fake news. >> thank you. >> and a very similar answer for us. i would only add that one of twitters great advantages in the world is that it's fast. it's faster than television news often. we try to arm emergency responders with the knowledge of how to use that as the strength so it's one of our key pieces during the hurricanes on the gulf coast, we were actively working with folks who were responding. there were actually folks in texas and houston who were using our platform to identify people to rescue. and so it's one of the strengths of our platform. and like everyone, it's a continuing challenge to address misinformation. >> mr. watts, would you like to address this or is it anything
3:13 pm
else that can be done? >> i don't know terms of the technical thing that can be done i do think the spread of misinformation so quickly like that, the first thing that you see is what you can do believe over time, that would use the most is what you can to please us will. it really empowers social media manipulators if you can do in plinks. if you can generate other system to push that is quicker than anybody else. you see a lot of gaming in terms of training hashtags and things like that. i think there has to be some sort of trip that you can put in technically over time, and i'm sure all these companies are trying to develop that will tap that out. when using artificial spike in any one of the strings you should be able to detect it. i think they are dancing on that. it's a huge public safety issue regardless of the threat after that is deploying them. >> thank you, and i know i'm running out of time to let me just say this. i had the opportunity to work with facebook on our internet crimes against children task force in nevada when i was attorney general. and i will tell you, for every
3:14 pm
company that we reached out to, whether it was youtube, google, they were willing to work with law enforcement. so window that hasn't been a lot of discussion on that interaction that you had with law enforcement but but i seenm one side of it. i know now there is this balance we need to find to figure out how to continue to work together to address these evolving crimes and activity that's happening on the internet. i'm grateful that you hear, and i look forward to figure out how we can continue to evolve that relationship as well, so thank you. >> thank you, senator cortes master. senator lee. >> thank you, mr. chairman. thanks issued you for being here. we live in an exciting world, in a time when companies represent at this table today, 15 years ago were just ideas, and today they change the way we interact with the world around us. today these companies have made
3:15 pm
it possible in ways never imagined just a couple of decades ago for few people with very few money to have an impact, not all in their community, but across the country and throughout the world. but with that come a lot of challenges, and those challenges are the reason why we are here today at this hearing. in some parts of the world there has been a suggestion that, i can summarize only as an effort to make public utility companies out of social media enterprises that would rather comprehensively attempt to regulate social media, and impog escalating fines and other penalties on companies that fail to report certain types of information to the government. some of these recommendations for policies like this at been made in the united kingdom and in the european union. to me this kind of distressing,
3:16 pm
in part because i worry about what that would do to private property, what that would do to these thriving businesses that have given so many people so much of an opportunity to be heard. i also worry about what it would do to public safety. the very end sought to be achieved by these proposals. sometimes when government gets involved, and sets a certain standard in place, that becomes both a floor in the ceiling, understandably. i would worry about that. we'll start with you, ms. bickert, and tell you what you think about proposals like that and what some of the risks might be to starting to treat social media companies like public utilities. >> thank you, senator. i think whenever we think about regulation, there often are
3:17 pm
unforeseen consequences, and those can impede our ability to provide services to the people that trust and need our products. i think the big thing for us is that our incentives are often aligned with those of government in terms of creating a safe community. on this issue absolutely. there is no question that the companies here do not want terrorist using their platforms. the long-term business interest for facebook is we need people to love a good experience when they come to facebook. we need them to like this community and the a part of it, and that means keeping them safe and removing that content. so the incentives are there. these companies are working together to address these challenges, and that's how we think it worked best. that said, will continue to toe a productive dialogue with government, the concerns that you face and what you hearing from your constituents matter to us very much and we want to make sure we're considering that and
3:18 pm
are responding to that. >> in light of the factor company and others, the progress that your company and others have made in this area doesn't suggest some of these proposals are unnecessary, in any event? >> thank you, senator. because our incentives are aligned, the kind of progress that you're going to see is going to happen regardless of what we are seeing from governments, what we are hearing from governments. it's still important to have that dialogue. we learn every time that we engage with policymakers, but the incentives exist independently. >> ms. downs, would you agree with that? >> yes. the security and integrity of our products is core to our business model, that includes the speediest enforcement of all our content policy. so we're already motivated and is this an essay resources and people and addressing this threat. >> how my treating you more like the public utility change that dynamic? >> i think the risks that you
3:19 pm
outlined are important things for policymakers to remain cognizant of, obviously the tech industry is incredibly innovative, has created tremendous economic opportunity, and anything that slows down that innovation will cause damage to the ability of the industry to continue to thrive. >> mr. monje? >> i would agree with that bit we take our responsibility extremely seriously, and it is important to our business to get it right. we measure progress in matters of weeks and months. we move very quickly, , so i wod agree with everything that was said. >> i've got one second remaining, if i can just -- how do you determine, will just go with you, mr. monje consent are already on you, how to determine what constitutes terrorists and extremists content? for example, do you make this determination internally within
3:20 pm
your staff? df certain subject matter experts that help you decide that. >> was yes, sir. with former enforcement officials were on our team. we also interact with and communicate with governments and ngos to determine that on an individual basis. >> okay. my times expected thank you very much. thank you, mr. chairman. >> thank you, senator lee. senator hassan. >> thank you very mu ch, mr. chairman pekin thank you to our panelists today. i want to talk about the "see something, say something" campaign. the campaign is simple, and many of the thwarted terrorist attacks in u.s. were stop because everyday people alerted authorities to suspicious behavior. so i'd like to get a better sense of whether your company fully embrace this "see something, say something" campaign. while i understand the most of you shut down accounts that
3:21 pm
espoused while extremist propaganda, it's not clear that you proactively report those accounts to law enforcement. here's an example of why that makes us less safe. in 2012, tamblyn tsarnaev, one of the boston marathon bombers posted on youtube several videos espousing al-qaeda propaganda under the name muaz. the fbi had present guest cameron thanks to tip from the russians the fun nothing to corroborate the russians claim. in september 2012, he applied for u.s. citizenship dhs. as part of the vetting process dhs instructed the fbi to run a check on the application which came up all clear. however, in his application, he revealed that he tried to change his legal name to muaz, the same name as a youtube account. eight months later he
3:22 pm
orchestrated a terrorist attack with his brother that result in the deaths of four people, and almost 300 injured. in hindsight, if you two have reported his troubling social media account to the fbi, and maybe get the i would've been able to link tamblyn to muaz extremist youtube account when hammerman was applying for citizenship. that could a profit the fbi to reopen a close terrorism investigation just weeks before he carried out his awful, tragic bombing. so the mr. monje and ms. bickert and ms. downs, i'd like to understand how and when your companies report extremist accounts to law enforcement. and has a change since the days of the boston marathon bombing? >> thank you, senator, for the question. when we are aware of an imminent threat we actually do proactively reached out to law enforcement. whenever they come to us and ask the right information when you
3:23 pm
have the right process, which were good with working out we respond as quickly as we can. >> thank you sigrid farewells to cooperate with law enforcement pursuant to valid legal process including emergency disclosure provisions or if we detect any content on our services that poses a threat to life, we proactively reported to law enforcement. >> ms. bickert? >> thank you saturday. the same answer. >> i thank you. i will say this "see something, say something" campaign is premise of something a little different than what you'd all said, because it's premised on if you think, not does this mean my definition of imminent danger. but we ask members of the public if they see something suspicious to step up. what you are all saying is if it meets certain criteria or if you're asked to come to think that's a little bit different so let me follow up with mr. watts. as as a former federal law enforcement officer how would you grade these companies performance and addressing
3:24 pm
violent extremist accounts? to you think they can do more to actively support federal law enforcement and counterterrorism officials? >> over the last decade or so, they've all done better. facebook and google have outpaced twitter. twitter, in my opinion, relies too much on technical signatures and doesn't staff on the threat intelligence level to the extent that they should. >> thank you very much. and thank you. that's all the question i have, mr. chair. >> senator peters. >> thank you, mr. chairman. and to each other what is is thank you being here today. an important topic and appreciate your active involvement in this. my question concerns the extent that algorithms are used and play a role in the problem, and how algorithms can also be used as a solution to this problem we are dealing with. i was pleased to read ms. bickert post in facebook newsroom that facebook has started using artificial intelligence to help counter
3:25 pm
terrorist threat at first on your platforms. the speed and breadth of the internet certainly makes it nearly impossible for humans to keep track of all this so we need to have a assistance to do that and the need to continually evolve if we're going to be effective in using them. however, it is likely that algorithms may be partly responsible for getting extremist material in front of users, whether it be in search results through facebook newsfeed for you tubes up next list or elsewhere. so my question is these algorithms are under your direct control as all platform providers can control the. what you specifically tuned to learn more about whether and how algorithms may be promoting extremist content? >> thank you, senator. the first thing that we need to do is make sure we are removing
3:26 pm
the terror content and then it doesn't matter here point . yoe it out of the equation, then the algorithm has no role in promoting it because the content is simply not available on facebook. that's something we do, as you pointed out, using technology to find the content. but we don't stop there. after we find an account that is associate with terrorism, the we remove that account we also stand up for medi-cal. we look at associate content, associate accounts and remove those as well. if we can get better in that space, then we can make sure the content is not appearing before our committee. >> ms. downs? >> thank you, senator. absolutely correct that the first priority is making sure that none of this content is on the platform in the first place. at the same time we also have teams that are protecting our algorithms from being gamed. this is a threat to our services and to her users experience on our services across many issues
3:27 pm
as we dedicated teams to make sure that people are not manipulate our systems and that their working as intended to serve relevant information to users who come to youtube. >> very similar answer from twitter. we have been able to use our machine learning, our algorithms to help identify more than 90% of the terrorist content that we taken down before anybody else brings it to our attention. 75% 75% those before they get to tweet once. also we protect our trends against manipulation. we've done that since 2014 and we continually improve our processes to protect our user experience. >> mr. watts? >> i would just note that any sort of algorithmic detection techniques is only as good of what's already been seen in the world which is part of the reason why the russians have been more successful in terms of social media manipulation. they understand the terms of service and have the capabilities to beat those systems, and they play within the rules.
3:28 pm
the smarter, better resourced, higher computational people around the world that want to use it was the better. it's got a bike zero day viruses in cybersecurity speak. cybersecurity protections, antivirus is only as good as what has been seen for in terms of malware. the only way to get in front of that is to combine really smart threat analyst on whatever threat act it is that's out there with the technologists, and his companies that do that do better in terms of getting a foot of these actions. >> what's your assessment of the companies here and others in the united states? >> yeah, i think facebook and google i've seen massive increases and much more success in that space. i think twitter gets beat oftentimes picnicking continue to get beat because they rely too heavily on technology. i don't think the partnerships they need to adequately get out in front of them. >> if i could respond to that because, you said it twice and it disagree. i think there are many external researchers who have said that a
3:29 pm
lot of this terrorist content doesn't come has moved off of our platform. the average isis account in 2014 at 177 followers. now they have 14. they measure their life on twitter in minutes and hours. we are extremely effective for taking them out. we do have the resources in place of the technology in place to fight. >> mr. watts? >> they get beat by a new terrorist group every few years. i mean, al-shabaab, to watch the entire attack go down on twitter, monitoring if it we achieve monitoring list that we watched all that. with isis and al-nusra in syria were able to build up lists of anywhere from three to 4000 terrorist terrorist accounting any given time. they do better after the fact once they pick up what the signatures are but the problem is you are always treated with the threat actor is. you are not sting out of fun of it which is why in cybersecurity space or even some social media company are taking this on now if you employ the threat analyst
3:30 pm
to work with the technologists otherwise they are always behind the curve. that way until the group creates enough signatures that they can detect it and then they can we did that. they are getting better all the time but ai and machine learning even with its advancement can only detect what's been seen before. what humans are very good at at least up until now and until they become autonomous machines out there, they are good at gaming systems and figure out ways around it. i think in the case of the russians, for example, and i seen the takedowns of their accounts by twitter, and they are hardly making a dent in what i'm seeing in terms of flows. i can't confirm all of those accounts that are out there, but i hear about troll farms. why do we think there's only one? i think in terms of moving for this got to be a much bigger focus for the social media companies. i know both analyst doug on facebook and google in that space, and some have been there longer than others but i think that's the right approach moving forward. >> thank you. i'm out of time.
3:31 pm
>> senator cruz. >> thank you, mr. chairman welcome to each of the witnesses. i'd like to start by asking each of the company representatives a simple question, which is the you consider your companies to be neutral public fora? ms. bickert? >> thank you, senator. the mission of our company is to connect people. we do not look at ideology or politics if it we want people e able to connect and share who they are. >> i'm just looking for a yes or no, when you consider yourself to be a neutral public forum. >> we do not have any policies about political ideology that affect our platform. >> ms. downs? >> yes. our goal is to design products for everyone come subject are policies and the limitations imposed on the content that people make sure our products. >> so you're saying you to consider youtube to be a a neul public for? >> creek.
3:32 pm
certain things are for ticketed by her committee guidelines which are spelled out and provide a public you to all of our users. >> mr. monje? >> yes, sir. >> let me focus for a minute, mr. monje come you know there have been several videos that were released in recent weeks that i and love of the people find troubling. want to give you an opportunity to respond to them. one individual described as a former twitter software engineer was captured on video saying the following, quote, one strategy is to shadow man have ultimate control. the idea of a shadow van is you been summoned but they don't know they been banned because they keep posting a no-win cease their content. so just think that no one is engaging with the content when in reality no one is seeing it. is that the practice that occurs at twitter? >> no, sir, we do not shadow van
3:33 pm
users. >> why would this individual described as a former twitter software engineer say that? >> thank you for the opportunity to respond, , senator about thi. these folks are caught on video. they were not speak on behalf of the company. are speaking and their capacity. we do not shadow ban folks here but we do do is, if an account is spamming, meaning engage in malicious automation, we will hide, make it harder for them to be found on a platform. if i could continue. that was one of the reasons why the efforts that we saw with the russian misinformation didn't hit as they can work as they were hoping for. we were able to stop it in real-time. >> another individual, twitter content review agent was quoted on video as saying, on stuff like that it was more discretion on your viewpoint. i guess i felt about a particular matter. yeah, if they said this is, quote, pro-trump, i don't don't
3:34 pm
want because it offends me, this, that the and i said i been the whole thing. goes over here and there like you know what, i don't like it. he's right, let's go, let's carry on. what's next? is that individual describing a practice that occurs at twitter? >> no, sir. we use algorithms as a way to come if we see an account as being abusive, that also will be down rate if they're there engn targeted abuse against minorities, if they are being, if they're consistently violating our terms of service but they haven't crossed the line into being suspended. we will make it less visible. what we won't do is make your followers always be able to see you, and we are not come we ensure that if you go on twitter at any moment speed i want to make sure i'm understanding you write that you are saying for some people who are posting your restrictive jewish only to those are actively following the? >> if we believe that they are
3:35 pm
engaged in malicious automation. if we believe that they're violating our terms of service when it comes to abuse. >> so is it your position that the individuals that are subject to this form of censorship are extremists or fringe? is that what you are telling us? >> it depends on the user. i can tell you that this is not something that we hide from the public. this is out in the open, the fact that we will reduce the visibility of tweets that are abusive or that are engaged in malicious automation. >> let me ask what about congressman marsha blackburn? issue someone you would consider somehow abusive or fringe or otherwise? >> no, sir. >> well then why did twitter restrict and censure her announcement video announcing as a candidate for united states senate? >> i want to be very clear about that, i think you for the question, is that we never removed her tweet.
3:36 pm
and what she did do is advertise on a platform. we do like many platforms have a higher standard when it comes to advertising because we are putting in front of people things they didn't ask to see. her video was reported to us. there was a decision that was made that was later reversed because of some of the language that was used in the account. it was a mistake and we acknowledged it. >> so her announcement was censored because it was pro-life content. has twitter ever centered anyone for pro-choice condit? >> she was never censored. >> so you're saying nothing happen to her tweet? >> her tweet got a lot of attention on the organic side. we action on accounts and we take our terms of service very safely pick sometimes we make the wrong decision. with action on all sides of issues and we strive to be better every day. >> let me ask a final question
3:37 pm
because my time has expired. ms. downs, i'd like to know, what is youtube's policy with respect to prager university and allegations that the content -- prager university is putting out or being restricted and centered by you to? >> as i mentioned we enforce our policies and it politically neutral way. in terms of the specifics of prager university is the subject of ongoing litigation so not free to comment on the specifics of that case. >> i will say the pattern of political censorship that we are seeing a cost that technology companies is highly concerning, and the opening question i asked you whether you're a neutral public forum, if you are a neutral public forum, that does not allow for political editorializing and censorship if you're not a neutral public forum, the entire predicate for liability immunity under the cda is claiming to be a neutral public form. so you see you can't have it both ways. thank you.
3:38 pm
>> thank you, senator cruz. i think we got exhausted all the questions. thank you all for being here. i think it's been a very informative session, and we all know that the internet is an incredibly powerful tool, which has, offers enormous benefits to people globally. but we also realize with it in a dangerous world and that there are people out there who want to do harm and bad things and are looking for any means in order to accomplish of those. and, of course, we know that in the modern world, cyber has become increasingly a tool of choice for a lot of bad actors. so we appreciate your informing us about steps that you're taking to try and police some of that bad behavior. as i said earlier we have constitutional protections in a
3:39 pm
bill of rights, and we also have, i think we want to make sure we have a light touch when it comes to regulating the internet, and that's certainly something about a hope that this committee will continue to support and that those at regulatory agencies will adopt as well. but we also want to make sure that we are doing what we can to keep our country safe. and so we appreciate the efforts that you undertaken already and those, as you continue to develop and look at ways to combat some of these threats that we face, and we hope that working together as partners, that we can do a better job. and there's always room for improvement. so thank you for what you've done and what you continue to do, , and we will look forward o discussing i'm sure in the future as the threats continue to evolve, things that we can do better.
3:40 pm
so thank you all for being here. i'm going to just say that before we close i've got a letter from the consumer extremism project highlighted its work on combating radicalization online, and i'm going to enter that into the record, also entered a piece by the "wall street journal" authored by the counter extremism project senior advisor doctor -- let me say this right -- underscored his work on this important issue. we'll keep hearing record open for a couple of weeks. senators are encouraged to submit any questions that they have for the record, and upon receipt of those questions we ask our witnesses to submit the written responses to the committee as quickly as possible. thank you all for being here. this hearing is adjourned. [inaudible conversations] [inaudible conversations]
3:41 pm
[inaudible conversations] [inaudible conversations] [inaudible conversations] [inaudible conversations] >> center, to question. where are you guys with --
3:42 pm
[inaudible] >> she still has not withdrawn her objection to the bill moving forward. we think that the compelling logic of saving some of the 35,000 lives that are killed every year on our highways is something that we ought to have members on both sides in support of, but at this point we don't believe senator feinstein will be on board. we hope she will get there. >> has she asked for anything in particular? >> i don't believe she's as anything in particular. there were a couple other objections on the democrat side, but we're willing to work with the people who have objections and address the concerns, if he can be done in a way that doesn't undermine the purpose and the basic framework of the legislation. >> as a follow-up, when you expect to have a bill in the senate? >> well, preferably when we have some democrats are willing to
3:43 pm
work with us. >> thank you. >> at the moment we don't have the willing partner. i hope we will. >> this was a wide-ranging hearing focusing on terrorism and the election, domestic terrorism, foreign terrorism. what are the next steps for the commerce committee kind of engaging with these companies and getting into either reform the practices or inform consumers better if they been exposed to foreign content? what you think needs to happen next? >> we will digest what we know today. i think this was a really good first step and it's consistent with our oversight role. i feel like the companies buying large were pretty responsive and i think we got sense for things they are already doing. i don't know at this point that it requires or necessitate any additional action but will continue to dialogue with them to keep up to speed on what they are doing and to determine if there are additional steps that the committee ought to take going forward, but at this point it really was more than anything
3:44 pm
else we know how important these platforms are to extremist groups, to recruit and radicalize folks that will commit violent acts against americans. and so we just want to make sure we're staying on top of that issue and staying on top of what these companies who have such powerful platforms are doing to prevent that kind of activity. >> would you say getting to the bottom of that, that terrorism recruiting thing and having that kind of content on platforms is more the aim of the commerce committee convening these hearings rather than trying to get forget what happened during the election? >> clearly our purpose in this, and you're right, anytime you open appearing like this it in something a wide ranging discussion. people got us questions on range of topics, but the purpose of the hearing, stated purpose, is to determine what steps, what role soul to meet platforms can play in helping to prevent future terrorist attacks here in the united states and around the world.
3:45 pm
and so the degree to which other members sort of geared off or deviated from that, that's going to happen anytime you have a hearing that specifically is what we're trying to accomplish. >> can we talked briefly about broadband infrastructure? i know the house has a series of resolutions and builds in you want to do here in january. can you sketch out a all a bitf what you're thinking of both the substance and the time when? >> we will have a hearing as soon as we have something from the administration and we expect that to happen sometime around the state of the union. we hope that the infrastructure proposal that's put forward by the administration addresses not only transportation projects, roads and bridges, but also broadband and a lot of the issues that are also under this jurisdiction. we will look forward to receiving that. we had a lot of conversations already with the administration. we've given them a lot of ideas about things that they could do
3:46 pm
to promote not only highways, transportation, aviation, railroad infrastructure, but also broadband infrastructure. >> back to the hearing, are you satisfied with what you heard from the companies today in terms of their cooperation with law enforcement, with the intelligence community? >> generally. i think they are understandably somewhat guarded in what they can say publicly because they know that the terrorists are watching. but i thought in light of what we are trying to accomplish with this hearing, i thought the responses were good. they were informative and that gave us better insight into not only sort of what they've done previously, but the steps that there currently taking to become more sophisticated in responding to try to prevent some of these terrorist activities on their platforms. so i felt it was a good hearing. >> you sit on net neutrality
3:47 pm
you're working on a bill? >> we have a draft bill that was out there two years ago which is still, i think, a good starting point in terms of legislation. it is a starting point, but we need democrats to work with us. the problem right now is that the cra, as you heard senator markey point out, it's something that's sort of a shiny object for democrats to shoot at but it doesn't get them the result. it's not going to pass the house. it's not going to be signed into law by the president. the best way to get a result if you want to address the issue of having in place what i think are generally agreed-upon principles for open internet with sensible guardrails around that in terms of getting too heavy on the regulatory side, the way to do that is through legislation. so our argument has been all along, democrats come to the table, let's solve this. i'm hoping that will happen but
3:48 pm
i don't think that the cra is, certainly isn't conducive to getting democrats to the table so will have to see who -- >> we got to get the box to his next meeting. >> you're not working on your own social media bill, right? and this issue advocacy ad sponsored disclosure, you are probably not for the come right? >> that bill has some support, as you know, but sure, will look at anything like that. that's kind of more logical activity i guess. it's maybe not as much under our committees of jurisdiction. but if there is a portal there that draws us into it, we'll pay attention to it. thank you. >> thanks, senator. >> thanks, guys. [inaudible conversations]
3:49 pm
[inaudible conversations] [inaudible conversations] [inaudible conversations] >> concerts and republicans are holding a policy retreat for the next three days at a west virginia report sort. vice presidents will be speaking live tonight at 7:15 p.m. here on c-span2. tomorrow it's president trump's turned to address the gop policy
3:50 pm
retreat. live coverage of his remarks begins at 12:45 p.m. eastern and that will be on c-span. for nearly 20 years in depth on booktv has featured the nation's best-known nonfiction writers for life conversations about their books. this year as a special project we're featuring best-selling fiction writers monthly program "in depth" fiction edition. join us live sunday at noon eastern with colson whitehead author of the 2016 best-selling novel the underground railroad which was awarded the pulitzer prize and the national book award. his other novels include zone one and the intuitionist. our special series "in depth" fiction edition with colson whitehead sunday live from noon to 3 p.m. eastern on booktv on c-span2. >> sunday night on "after
3:51 pm
words," former speechwriter for president george w. bush and atlantic columnist david frum with this book trump walker see the corruption of the american republic piggies interviewed by "washington post" nonfiction book critic. >> it come comes from the samet as democracy and autocracy is a book about the state of power. that's what the suffix means. this is the study of donald trump's power, how did he get it, had to see get away with it. so it is the system of enabling pickets assistant in the white house and the schism between trump and congress, the schism between trump and the media that enabled them and create an audience, the system that involves the republicans don't really come traditional elements of the republican party and above all between him and that core group within the republican party who enabled him to win the republican nomination and then go on to the presidency. >> watch "after words" sunday night at 9 p.m. eastern on
3:52 pm
c-span2's booktv. >> legislators regulars and tech executives talk about cybersecurity, online advertising, digital innovation and more, all a part of an internet policy conference held earlier this week. the internet education foundation host of the event. >> hello. welcome, everybody. welcome. hello, welcome to everybody. welcome to the 14th annual state of the net conference. thank you all for coming. some of you have been here for 14 years in

27 Views

info Stream Only

Uploaded by TV Archive on