Skip to main content

tv   George Washington University Discussion on Encryption Technology and...  CSPAN  June 7, 2019 1:08pm-2:10pm EDT

1:08 pm
embodies, as margaret thatcher said, it is made by a philosophy that is right. it is not suitable for all people at all times but everyone ought to aspire to it. so i set up a nationalist. i don't want to exploit it. i want to make it available to people. i want to help them where we can we have a lot of experience with the civil society, a democratic society. am a mild nationalist. >> sunday at 9:00 p.m. eastern on book tv on c-span2. found ice'sstudy supporters are using encryption programs to communicate with supporters around the world and to recruit people by using programs like telegram. we will hear from the authors of the report put up by george washington university's program on extremism.
1:09 pm
talking] >> good morning. thank you for coming. on behalf of our director and myself in the program on a stream is him, welcome to george washington university. we are going to launch a report to years in the making entitled "encrypted streams him: -- extremism." this is ays culmination of two years of research with a cooperation with the gelman library and the scholar technology group. researchers to my left have been collecting and coding more than 600 isis telegram channels for about a year and a half. we will dive into the report's findings, which are now available on our website. to our twitter followers,, welcome. to c-span, things are watching.
1:10 pm
the first panelist will be bennett, a research fellow with the program. he talk about the first half of the report, followed by helen kristi powell, a presidential fellow who has been with us for the last two years to talk about the remaining parts of the report. we will end with nick rasmussen, former national counterterrorism director, and my old boss actually. senior director at the mccain institute for national security and counterterrorism. we will do probably 30 minutes of presentation and then moving to question and answer. they will be roaming mics. if you have questions, feel free to reach out. thank you. bennett? >> thank you. with any big report release like this some thanks are in order. to start off the first will go to the entirety of the program on extremism. staff starting from the top down with the director, lorenzo , who wasnd shamus
1:11 pm
supportive of the project since the date of inception and helped us out with various components. the rest of the program staff, especially audrey alexander, could not be with us today but was into middle and helping us design the report, collected data for the report and helping out also with the finished product, and our other colleagues john lewis, andrew, who gave insight. outside of the realm of the program on a stream is him, a number of external experts provided insightful edits on the rest of the report, which we are grateful for and also thankful for the scholarly technology group and the gelman library at basty well as clarissa helping out with the data analysis tool in the final design. i will be presenting the background behind this report and the project behind it as
1:12 pm
well as some of the major findings. then i will turn the floor over to my wonderful co-author, helen powell to present some of the case studies we did, as well as some of the recommendations and conclusions. folks that areor less familiar with telegram, it's an online instant messenger. if you have used whatsapp or signal, is very similar. there are three major differences important in looking at why terrorist groups and the islamic state make use of it. the first is there are a variety of communications options. everything from one-to-one secret chats and voice calls between two or a handful of accounts, and groups in super groups that can include up to 200,000 members. there are channels which have one person that broadcast to a theoretically unlimited number of followers. in addition to the communications options there is internal filesharing. you can post an unlimited number
1:13 pm
of documents, audio files and voice recordings of up to 1.5 gigabytes of memory, twice the afforded offered by other major social media companies and messaging applications. the one thing, they brought up when it comes to telegram is the privacy option. communication has different levels of encryption but potentially, more importantly each channel or group can have a private option, meaning it can only be accessed it searchablend within telegram but not on the public web. a pledgeegram offers to not share any information relating to user data with third parties, including governments. the current assessment is telegram is the preferred online platform for the islamic state supporters. a lot of this came in the wake of policies employed by major social media companies, twitter, facebook, google and the like to
1:14 pm
push islamic state supporters, their accounts and content off those platforms in the wake of a lot of these features, isis supporters concentrated many more of their online efforts towards telegram. thatimportant to note telegram has always been used by the islamic state's online supporters, it's just a question to the degree. it's always been a first step in this multiplatform distribution effort to push propaganda from telegram onto mainstream social media sites. was usedusly, telegram in the guidance of several isis -directed attacks throughout western europe and north america. 2015famously in paris in and brussels in 2016. telegram has a pledge to not share user information with third parties, hitting governments. regardingsions
1:15 pm
content regulation are little different than most mainstream social media providers. they will regulate any public content. public channels. but they will not regulate private content, including private channels, groups, or voice chats and secret calls. any of those options are not subject to moderation. in august 2018 in response to the eu regulations, they said they would "if it receives a court, order we may disclose your ip address and phone number to the relevant authorities." they note that to date they have not taken that enforcement action. if they do, they will publish it in a transparency report. to be brief about this program or project's methodology, there is a very full description of the process we used in the full report. between june 2017 and october 2018, we directed a team of
1:16 pm
program on a stream is a research assistance. we used three anonymous accounts to access, joint and collect pdf readouts of 600 36 pro islamic state telegram groups and channels, withing the sign which content. those are included with the final report. classifiedpdfs we content off of them using a number of quantitative and qualitative variables. working with the pipe bomb tool developed at the george the urlon university, key is used to access channels and groups, hashtags, and external links to websites outside of telegram. these were the main research questions in looking at the data. how do english speaking supporters use telegram to build online networks, disseminate propaganda and guide operations? secondly, in which ways do english speaking supporters balance the need for broad-based messaging and recruitment what
1:17 pm
the necessity of operational security? third, how do english speaking supporters on telegram reactive pressure against the organization in the online and off-line spaces? the following sections of the presentation i will present our findings. the first question we looked at one.e basic how and why are islamic state supporters using telegram versus other platforms? which specific features are most important to them? our analysis finds generally speaking these channels and groups can be classified into five major categories. throughout this sample, three major tactics are used. the first are the major categories that you can see on the graph on the left. the most important was distribution channels and groups. these represent a majority of the sample. these are used to distribute pro propaganda -- pro-isis
1:18 pm
propaganda. you will see the distinction made between official and unofficial media, sources directed by the islamic state central leadership and those that are not. in these distribution channels there is usually a blend of those type of sources. versus the second most popular option, core channel and groups that only disturbed it official propaganda and disputed propaganda not produced by the islamic state central media division. on the bottom there is another notable category, instructional material. this is the compiled set of constructions for assisting terrorist groups conspicuously. it includes everything from bomb making materials to cybersecurity manuals. helen will cover this in her section. categories, we see three major tactics illustrated in these three graphs. on top we have helped islamic state supporters use telegram's internal filesharing features.
1:19 pm
there is a vast amount of information contained within this sample. over 63,000 photos, 22,000 documents, 10,000 videos, 4000 audio files and 900 voice messages. in addition to internal filesharing, we have a network graph of links. the ones in blue represent joint links shared in between channels and groups within our sample. the ones leading to the external node where shared external to the sample. this demonstrates the endogeneity of the sample and maintains the degree of connection to which these groups and channels are all connected. they are using them to essentially add new users, make sure any users within the sample is joining while the new group -- ensuring the resiliency of the islamic state network on telegram. on the bottom right, it's difficult to read, a set of instructions that preceded a
1:20 pm
number of downloads for cybersecurity-related programs. this is the third major tactic, just riveting a lot of information pertaining to operational and cyber's ready specifically -- cybersecurity specifically. moving on to the next finding. this is sort of what we have termed the online extremist dilemma and using digital communications technology. at the same time islamic state supporters must both make sure they are reaching out to the mass public to ensure their narrative is injected within popular discourse, but also they are trying to protect their network and ensure everybody is using encryption and ensuring everyone is using privacy maximizing services in order to protect their network from external infiltration and takedown. saw is these two goals are fundamentally at conflict with each other. on one hand islamic state
1:21 pm
supporters had a strong emphasis on privacy. that is demonstrated to the graph on the left which shows the majority of our sample was comprised of private channels. as well as if you add the group features, that's an additional 10%. it comprises about 70% of the sample that is not going to be regulated by telegram under its current policy. islamic state supporters are deliberately choosing these groups to ensure their content and networks estate up in the face of takedowns and protective information from external viewers. the difficulty of this is that using these close groups and quorums hamstring the ability to recruit. our analysis on the right is over 40,000 urls and destinations outside of telegram. these are the efforts islamic state supporters are using to ensure their message is broadcasting to social media sites like twitter, youtube and google, and to filesharing
1:22 pm
platforms included in there as well. this messes with operational security. external filesharing and posting on social media. one supporters do that they potentially release information that can provide a link between their telegram account and their presence elsewhere online such as an ip address. unfortunately this is ripe for exploitation -- fortunately. unfortunately for islamic state. supporters --islamic state supporters. law enforcement has built investigations of the islamic state supporters. moving on to the final finding. this is how supporters are reacting to the massive degree of loss they incurred during the time period of the sample. it covers the time the islamic state lost its major stronghold in syria and iraq. other defeats,
1:23 pm
and massive pressure by social media companies against the islamic state on the public web. how are supporters reacting? we found they are refocusing on the islamic state's military activities, attempting to ensure resilience of their networks online, supplementing the islamic state's official propaganda with user-produced content that follows the branding guidelines and the makeup of islamic state media without being directed produced -- directly produced by central media apparatus. in developing new measures for the online guidance of operations. we use a couple of metrics. the first is a qualitative analysis of content. the most discussed topics in the sample were the islamic state's military activities and battles in syria and iraq, as well as through its affiliates. one was that shtag graph.
1:24 pm
most of them are centered on events in syria and iraq. all of these are registering this top 25 hashtags and others that are outside, like the site of a siege from the islamic state's affiliate in the philippines. what is absent is any discussion of terrorist attacks in the west by name. that mirrors other studies having was speaking is supporters. they are engaged with things that are happening on the ground in syria and iraq, especially in light of the territorial collapse of the islamic state. i will turn the floor over to helen. this is sort of the top-level but we did individual case studies we observed. she will talk about our recommendations and conclusions. begin i must admit i
1:25 pm
have a couple of things to say. thank you for taking on as a presidential fellow in the fellowship program. i have been able to work at the program as a research fellow while doing my masters at gw. it's been a great opportunity where i have worked alongside other great researchers like in it and audrey. i'm excited to share our research on telegram, the culminating experience of my two-your fellowship. -- two-year fellowship. while we do have front and observations of telegram while collecting data for the past year and a half, while we were doing this three individuals were arrested in relation to their pro i.s. activities online. some activity was captured in our sample. while bennett presented our front end analyses and assessments of how we believe ice is are using telegram, these help us make a better picture of what is going on behind the
1:26 pm
screen and looking at the criminal completes to see -- complaints to see how they used telegram to achieve specific goals. we find the three case studies and body three themes going on right now. the first is a rising grassroots actors. the second is the proliferation of unofficial or gray media. the third is that this region of operational and instructional material. i will go through the studies now. the first is a case study of karen. she represents the rise of grassroots actors. she was arrested on october of 2017 and the philippines. the media knows her as recruiting, primarily folks from india, to come to the battle to fight with the islamic state. on telegram she talked about her -- she was talked about in a different light. a lot of people called her a spy. karen was able to build a very large international network.
1:27 pm
there are a lot of people on here but i will highlight a few. the first is foreign fighter at the top left. he traveled from india to syria. if you read the criminal complaint for mohammed nasser, it's connected to a larger cell of individuals arrested in india for attempting to start an affiliate. these criminal complaints coming from india are quite interesting and definitely worth a read. network and was pretty aggressive in her tactics to get to this point as a central node in the telegram ecosystem. she assumed a lot of authority and got a little careless. she was not following a lot of the typical norms you would see. people did not always like that. mohammed, another person she was connected to, started a feud with her because she was flippant about cyber security measures like protecting cell phone numbers.
1:28 pm
he started a movement against her, starting channels against karen. people began calling her a spy or a disbeliever and trying to push her out of the space. it goes to show while you can build a vast international network, there are still norms and protocols. to fit into the i.s. supporter ecosystem on telegram. , hesecond case study embodies the proliferation of unofficial gray media. he was arrested in october of 2018 in chicago. or is going to be on trial for material support for the islamic state. on telegram he was best connected to a part of anonymous media foundation, a gray media group. we observed a lot of these groups operating on telegram. we could only make our best assessments of what is going on
1:29 pm
behind the scenes. kms was one that we did observe. the criminal complaints details how the media foundation works, providing a vital window into how one specific example of a media group operated. the criminal complaint details how they intentionally replicated and duplicated i.s. content, taking new videos. this is something we often see in these groups, replicating and duplicating official content. they had a hierarchy with different publishing groups and editing divisions. they would pass products through these divisions before being published and translated into more public spaces. it shows how they used telegram to their advantage, where they would plan raids to inject into the public discourse on twitter. and having private leadership groups where decisions were made before letting the other staff groups know about things.
1:30 pm
one thing is that the complaint shows the media foundation has some connection to i.s. central leadership. this was demonstrated when the media try to start a merger with another media group. they started this merger and then actually rescinded it once someone from i.s. central leadership said don't do this. this goes to show an example of how these media groups can have some sort of connection and direction and check back in with a central media group when making decisions to get course corrections along the way. our third study example applies the continued and grassroots-led dissemination of material for guided operations. he was arrested in november 2017 and the u.k. he was known in the media for threatening prince george's school.
1:31 pm
astelegram we know him best sharing operational materials under the moniker -- his case helped us determine that instructional and operational material oddly fits into three categories. the first is explosive construction. the second's low-tech methods like knife attacks or truck ramming attacks. the third is operational security and cyber security protocols. to use a vpn or a fake phone number. most pro i.s. channels have a strict ideological adherence to material, the instruction materials were often native of any and all materials, including manuals produced by al qaeda or chemistry textbooks or screenshots of how to -- an example taken from google. another interesting thing about this case was the fact that the monikeriker -- lone
1:32 pm
posted after his arrest. this make indicate there were other co-conspirators working under that moniker before or other individuals began posting under that name after his arrest. anonymity provides is very strong. it can be difficult to know if multiple people are acting behind when username or if when username can have many more people taking up the flag. it can be a bit of a challenge when it comes to combating extremist exploitation of telegram. that, what do we do about that? we have five considerations coming from this report that we would like to share. the first is a large-scale massive shift away from telegram. we assess it is unlikely. the features are really great. it combines unique elements of social media where they can have a public broadcasting element,
1:33 pm
as well as encrypted messengers. this suite of features is useful. they previously tried to jump to other platforms that may be more secure. until there is a viable replacement that has a lot of --se combinations of public and encryption and user-friendliness, we assessed a current shift is unlikely. the marginalization paradigm should shape our approach when dealing with extremist expectation of telegram and online. the people we have been observing our fast at adapting. instead of trying to chase them from platform to platform, we think we should think smarter about how we can limit and confined extremism. policymakers in companies should follow the our generalization paradigm set up by a few other researchers in the field which tries to confined extremism to places where they can't inject
1:34 pm
into the public discourse and law enforcement can still monitor them. that brings us to the third consideration. i.s. our currently marginalized to telegram. accessible,publicly but it is still separated from mainstream social media. he can't get the telegram by google or facebook. you would be blocked by a prompt to login. the islamic state is continuously trying to conduct outreach to recruit new members. because of that window law enforcement are able to monitor. pushing them off the platform may not be the best plan of action. our fourth consideration is that government should encourage telegram to participate in industry-led forums. telegram does monitor and take down i.s. on the public channels.
1:35 pm
when they do this they see a lot of stuff going on. it would be great to have them share their insight and trends to help other public tech companies do their own content moderation. if telegram identifies a new gray media group, they could alert other companies like facebook and twitter so they can quickly tag and identify it on their own sites. another area for cooperation is alerting smaller companies when i.s. supporters are trying to jump to their platform. we are able to see where they are linking to. if telegram could give a heads up, smaller companies could say i.s. i.s. we are seeing a lot of supporters -- we are seeing a lot of i.s. linked to your website. they would not have to change their stance on privacy or get governments involved but enhance current practices in place by industry leaders. the last one is the heavy-handed
1:36 pm
approaches that a disproportionate and consequential. , while it may be tempting to say why do we just shut down telegram orban encryption -- or ban the encryption? iran and russia. russia tried to do it and almost shut down the entire russian internet because they blocked a lot of google and amazon ip servers. in iran they struck,. iran, they have struck a major blow to their economy. we can focus on these cooperative measures building off of existing practices to better define extremism on the platform. that concludes our brief overview of the report. it's available online and we look forward to your questions. >> great. i will jump in last. respect to what they just let me say something
1:37 pm
about the program on extreme is a more broadway. there is a sweet spot in this area. when you hit that sweet spot where you are both academically rigorous and policy relevant, you have really found the right spot. sorry about that. there are a select few institutions for that is happening in poe is one of those. i consider poe in the work of gw in this area as a go to destination for scholars and practitioners in the extremism and terrorism field. kudos for creating that architecture over the last several years. kudos to bennett and helen for producing this study. have genuinethis policy relevance. in that vein they help shape government policy.
1:38 pm
some of the conclusions helen highlighted can shape and inform government policy. i don't say this lately and the -- this is not always the case with work in the national security realm done in the academic world. there are plenty of areas, including in my own area of expertise, terrorism, where the advantage that government has over what is available information-wise is simply too great. and that the work done on the outside is interesting and informative but not necessarily in a position to shape government policy and the thinking of senior government officials and the weight i think academics hope when they produce studies of this sort. this study actually hits that mark and it does so because it is in some ways a comparative advantage that shifts in the direction of a place like poe and direction of scholars like bennett and helen. and thee the time
1:39 pm
wherewithal and the resources to step back and study a phenomenon deeply in a way that government analysts would not. anddelta between government the outside world in terms of data is pretty much nonexistent. they are looking at the same data that my colleagues and government would be looking at. that underscores the degree to which there is policy relevance to the study because poe can in fact do just as good a job of analyzing the data as i would argue anybody in government. selected from what they briefed to you over the last 20 minutes to me are first of all the eye-opening bit was the continued operational significance of telegram as a platform. this is not a platform simply used two recruit or spread messages or spread the word or perpetuate the narrative. all those things i would've taken as a given.
1:40 pm
what struck me was the continued operational relevance. the idea individuals could be engaged in activity that would lead to operational outcomes. the three case studies that helen highlighted underscore that in stark detail. if there were these case studies, there are certainly a number more buried in unavailable information. you can assume the scope and scale goes beyond just those three case studies. they cover significant theaters where isis tries work. united states, united kingdom and southeast asia. not just southeast asia but the first case study of the individual involved ahead to well beyond southeast asia. isis continues to carry out a global -- to represent a global enterprise. i think most scholars in the
1:41 pm
field would understand the physical defeat of the caliphate in iraq and syria does not change that reality. most scholars recognize that. it's important to bring that home to the public and to the broader policy community so that is not lost, even amidst the successes we have enjoyed against isis. i would pointway to and stop is the roadmap this provides for government and private sector engagement with telegram. beenlen noted, there have obviously attempts to work with the technology sector to address these challenges of terrorist use or extremist use of online platforms. those have come with mixed success. some success with some companies has not translated across the industry. studies like this one will arm government officials with useful data and information they can
1:42 pm
literally drop on the table and atre with senior leaders companies like telegram. we don't have to deal with the questions of what can be released from the government, what is classified and not classified. this is simply publicly available data case that this platform continues to be used in operational ways by terrorist organizations. that's a tremendous contribution from poe and the study authors. i will stop there. >> thank you. thank you for the kind words. they can be shared with the work you do at the buchanan institute. moderate hishe prerogative and open it up to the floor for larger question. struck by the section on the gray media. i remember the front page of the new york post with a picture on christmas of a beheading.
1:43 pm
that was created by a gray media organization. how important is this mix between the official and unofficial as isis continues to lose territory as the propaganda output has been lower and lower? is gray media the new normal for a little bit? helen: when we first started looking at telegram, that was one of the first things i was surprised by. the vast amount of gray media. you see a big spectrum coming from individuals and individual means where people are taken images and put new text or altered it in some small way, or mashup videos where they can combine elements from other media pieces to make new propaganda outlets and creations. studytried to conduct a and fill out that spectrum. it proved to be difficult.
1:44 pm
particularly on telegram you have the and out of position built in to the application -- ation built into the application. that is one of the big challenges we face when confronted with a lot of this gray media in figuring out who is producing it, where it's coming from and the impact it has. it is difficult to tell if there is an isis supporter differentiating between gray media sources and official media . it translates the whole spectrum of the gray media outlets. it provided an interesting window into the backend of what can be an example of what goes on within a gray media organization. it is something to look forward to and i hope other researchers will continue to study gray media and the impact it has.
1:45 pm
it will be here to stay. >> i would jump in and add something. insight government a tremendous amount of effort is devoted to mapping formal isis structures that produce media. who, what, when, where, why. to enable effective disruption operations where those were possible. that was always incredibly hard and incredibly important work. what helen just said reveals it is also the tip of an iceberg. what we don't know and have much less insight into is what goes on below the waterline with gray media. the academic world can probably help the government understand exactly how much is below the waterline and how much operational significance is happening below the waterline. nick may want to jump in on this. [indiscernible] enforcement will not have access to this information [indiscernible]
1:46 pm
can we talk a little bit about why they decided to go with telegram? what level of encryption are they dealing with? is it the right level? how has law enforcement work around that? >> the important thing here is that telegram offers a variety of different communication options. different communication options have different levels of encryption. some of them, especially secret chats and voice calls are protected by end-to-end encryption. i'm not a cryptographer myself but the protocol that telegram uses to engage has been lambasted by a lot of experts in cryptography. they call it homebrew encryption. i don't know about this and can't do it independently but what i will say is there are certain portions of telegram
1:47 pm
that definitely enable the going dark phenomenon. i remember this being a major concern after the paris and brussels attacks. police in the eu could not access the messages the various attackers were using to communicate amongst each other. as with regards to telegram, there are other parts that would not facilitate the going dark problem. in some cases this desire to want to reach out to platforms outside of their internal q negation can also be a benefit for those looking to infiltrate and disrupt their operations as well. at some point in time you will have to outreach to somebody else and you are taking the risk your parents tell you to never do on the internet. don't trust someone you don't know on the internet. that person could be a recruit or a detractor. there are some advantages or waste to get a window into that. the other downside for telegram for islamic state supporters is
1:48 pm
when somebody has one foot inside the general ecosystem through accessing other links or building a rapport with other individuals within the ecosystem, they can ensure they get access to think that would not necessarily be in the public eye otherwise. >> i will open it up to questions. please raise your hand and someone will come around with the mic. don't be shy. yes. hi. thank you so much for your research and how long it must've taken to get this information. i'm a specialist at ar international consulting. my main question is there are a lot of other intelligence you can gather from these messages. was there any common demographic information you found? whether it be what you showed earlier with the international reach or age, education,
1:49 pm
training? how much of it can be trusted versus not be trusted? >> i will take that one. things,he interesting and i believe we have a section on this, is whereas on twitter in 2015 you would see a lot of pro isis supporters -- supporters posting pictures of themselves. they would have their hometown in it. on telegram we are seeing a much more culture of operational security where the information is protected. that is what got karen in trouble. she was passing around his personally identifiable information. on telegram we don't see quite as much of that being passed around because they are so concerned with operational security. i am sure if you were able to talk to them you could get some of that but with our study we did not communicate with anyone. on the face level we are just not seeing that information shared.
1:50 pm
we will wait for the mic. sorry. >> thank you so much for this talk. my name is natalie hicks. i was wondering regarding how karen was out cast because of her -- outcast because of her misuse of the operational norms. what were more of those norms and could sewing the distrust and accusing people of not following those potentially weaken the network? >> yes. with karen, on your point of other examples, the main example is the sharing of phone numbers. this comes from the indian criminal complaint for mohammed. that is the one kind of verifiable fact that we have on the back end from that case. on telegram people were accusing her of many other things.
1:51 pm
of being connected to the arrests of a lot of other i.s. supporters. as far as having fact-based analysis and not coming from a primary source on telegram, the main example is the sharing of phone numbers. >> other questions? yes, ma'am. thank you all. my question is about the recent legislation proposed in the u.k. and passed in australia having a more heavy-handed approach to terrorist content on the internet. do you think policies like that will further the marginalization paradigm and restrict terrorist content sharing and recruitment to applications like telegram? or will it lead to telegram being ostracized in these countries and making, i don't
1:52 pm
know, then not use telegram and find other apps? how do you think that will influence terrorist use of telegram? >> you may want to jump in. one of the difficulties there is a very careful balance that needs to be played. one reason we sort of caution against the use approachs like fines against social media providers and heavy-handed approaches is it is difficult to target to a specific platform. one thing that is hard to do is it has to be applied throughout an entire sector. eventually any online website that does not follow guidelines put forward in the u.k. or australia is subject to the regulation. it's a good strategy if you're trying to implement the goal of getting terrorists of the internet altogether. that is not necessarily a tenable goal.
1:53 pm
one reason i think that we are dealing with this entire idea of isis on telegram as much as they are is because when the regulations were put into place against twitter and facebook it is very difficult to contain their impact it just goes platform. happening is up these large-scale migrations to platforms we have less of an indication about. --is soda following like sort of following cold air and hot air in a blink. if there is a lot of pressure, the migration will float to wherever regulations are not put in place. >> governments in australia and the u.k. are feeling genuine public pressure to transfer that pressure in the direction of technology sector. they don't have the ability as our government doesn't the u.s. to say the first amendment and use that as a shield.
1:54 pm
i take the point that bennett makes. this is some way pushing users to places where we have more difficulty learning and gaining value from what they're doing online. that has its own salutary value. if you are directing them to a smaller number of platforms, that makes governments' job to be like this platforms a little easier. obviously a cap on how much time, effort and energy engineers who work in places where the government worry about the problem can spend. if they're spending it on 15 platforms instead of five, that makes their job harder. i would argue it's a benefit. regulation -- radicalization pool may shrink. another question here. >> thanks. this was a report on
1:55 pm
english-speaking telegram. i'm wondering if there are any insights you gathered in your research or if you could pontificate on the bigger arabic picture or also french, german, russian, other language, gray and official media. thank you. >> the degree of interconnectedness between languages is important. it is part of the gray media picture as well. a lot of what these media outlets are doing is taking original content and producing their own content. it's an unofficial translation of officially released content. gapbviously takes out the when previously you would have the media center or another official outlet doing the bulk of the work and translating materials into these different languages. now it can be outsourced to supporters. the downside is if i less good of a translation.
1:56 pm
it may follow a less strict hierarchical rule. it can reach into languages or areas where the islamic state may not have had capacity before. the other thing we see, even though it is english language, we see a lot of intersection with the other types of languages shared. there is a decent amount of arabic language stuff. english is important as well in this context. it is the digital -- in the digital arena it is a global lingua franca, and a jihadist lingua franca as well. an isa supporter in the philippines, in nigeria, south america and washington, d.c. may not eight be able to speak to each other in arabic, but more likely in english. that is something our sample gets to a little bit. the english language does not mean it is a geographically restrained as well. >> can i adequate question. beyond the islamic state or i.s.
1:57 pm
problems here, did you pick up anything anecdotally or impressionistic late about the y aboutimpressionisticl the weight -- the focus of the study, obviously. >> we saw some cross action where we saw some of the insight. it was in disputes. someone went in and isis telegram group and forward something else from an al qaeda groups, telegram forward it into the channel and critique it. we got some window of insight into that. other detractors as well. the number of the shia militia groups in iraq and elsewhere are also active on telegram. there have been recent reports this is beyond the scope that certain far right it she was groups are also migrating to telegram. very similarly in the wake of
1:58 pm
increased pressure by major social media groups to push them off their platforms. that is something to consider even though it is not within the scope of our study. >> i'm from the australian embassy. i would like to echo nick's comments and commend you for the work you have undertaken here. not just in terms of relatives -- relevance but in the timeliness of the work. we heard about australia's experience when he came to legislating against airport terrorist-related material on the internet. timely because were undertaking this work with g7 partners. a lot of the push from our government relates to the events in christchurch some months ago. there were references to life streaming. i wonder if you had any reflections on the one in which
1:59 pm
-- the way in which life streaming is used on telegram. thank you. i can't remember. it was not something we look specifically at, if there was livestream video in a similar manner. but you saw an instantaneous reaction when there is a significant attack. the hashtag count i showed you before. and just generally about content. one of the reasons why attacks in the west or australia or anywhere outside of i.s. territory did not register either as a content category or top 25 hashtag is because that in isistopic tends to be hot telegram groups. you will have supporters that 40 engaged with it for 24,
2:00 pm
8, 96 hour time window after the event has actually taken place and after news media is released, but does not do anything much unlike a major battle or siege or event to drive the conversation over the course of our a longer time. that is what our general finding was in terms of instantaneous reactions to these sorts of events. question,r one more sir. ctc at west point recently released a study of one of its findings showing that essential media was very concerned with the centralization of the content that was produced, so i am wondering if in your research on gray media, do you have any insights on the relationship entialn official isis ess media and these unessential media accounts. >> to the rabbit hole, i love that. >> yes.
2:01 pm
my study, the criminal complaint shows some interesting kind of rules that were actually posted within the writers group i believe, for the media foundation. i am paraphrasing these rules or something like that, you always follow an official announcement claim only for announcements that were pro i.s. these rules kind of allowed a few guidelines to help the media foundation produce the of stayda, and kind with and a couple of guidelines. so i think, it demonstrated and need and did -- and then attempt to try to centralize and keep them somewhat in line, but direct oversight and approval for publication -- before publication was not there. within these guidelines, they
2:02 pm
were able to produce whatever mash-up that fit the bill. we did see a connection for the larger issues, so for example, the merger, but i think it was interesting that they had already begun the merger when they checked in with i.s. central headquarters and were told to resend. there are those kinds of the times within the media foundation case. unfortunately, without any other case studies, it is difficult to say what is going on behind the scenes, but the criminal complaint does show and demonstrate how they are attempting to do this in one example. >> can i add one more comment? this is based off of inference as well, but when i central media generally releases a statement that says do not do x, tois usually it is safe assume that at the grassroots level, someone is doing x. i think that is especially true in terms of the centralization question.
2:03 pm
gray media in response to the due centralization was posting content that was at times, divergent from what i.s. central media lines were. i can see in general, like a lot of these generalizations after reading the ctc study, a lot of the regulations came around the same time when there was significant disputes at the grassroots level of whether to follow the media outlets or follow i.s. central media in their line as well. >> one last question. hello, i am with chicago projects on security contracts. i want to thank helen and the team for their fascinating research. how isisy question is reclaims -- in 2018er of a video
2:04 pm
where the group very publicly acknowledges the fact that it's channels are being infiltrated online by western and u.s. security agencies, but the group's reaction to that is with every channel taken down on telegram, we will create 30 more. it also features supporters literally sending the pledge of allegiance online via text, so , ih this sort of reaction was curious as to what exactly are they talking about in terms of instructions on how to maintain operation security given that expansion. thank you. my mike is already turned on. categories the two which are posed under the same header our operational security generally -- that is just taken
2:05 pm
care of. most of that is an online. do not talk to other people that you met on telegram about trying to conduct an attack. be careful of other people in groups, and one of our offers is distributing a manual for basic operational security of how to avoid a tail, like someone driving behind you in a car, or to not drive through major roads that have ctc cameras on them. but the more important category i would say is cybersecurity, and that is to use privacy maximizing services. here, we have been infiltrated, and what you need to do is not just used telegram and its encryption, but here's instructions on how to download a vpn, here is how to use a secure email service. here's estimate sure you are not
2:06 pm
sharing other information on your telegram account, measures like that. i would say those two categories are the largest of what we are .eeing with isa supporters >> thank you. i am aware of our time and, and i want to remind you that the report is available on our website. i want to congratulate ben and helen for a job well done on a two-year project and it was a pleasure as a member of the team to watch you go through the process and i am happy that it was not me. coding, and that thank you, nick, very much for your insights on the report. these guys will stick around afterwards and if anybody has any questions, and the livestream for c-span, thank you for watching. [applause]
2:07 pm
>> watch live coverage of the 2020 democratic presidential candidates in iowa this weekend on c-span. -- tonightcease on c-span, we are with bernie sanders. and then cory booker in iowa city. democraticn sunday, presidential candidate with iowa speakersc party -- include pete buttigieg, kamala harris, amy klobuchar, beddoe o'rourke, and elizabeth warren. theh our live coverage of 2020 democratic candidates in iowa this weekend on c-span. watch anytime on or listen with the free c-span radio app. >> on monday, the house rules committee is expected to debate
2:08 pm
authorizing judiciary committee chair jerry nadler to pursue federal court enforcement for subpoenas for attorney general william barr, and white house counsel, don mcgann. watch live coverage on, or listen live with the free c-span radio app. sunday on q&a, blues musician daryl davis talks about his book "clandestine relationships" where he details befriending ku klux klan members. >> he was walking in first and he was wearing military and thege with fatigues initials kkk on his chest. it was embroidered across his beret, and members of the ku klux klan. hip, he had a semi-automatic handgun in a holster.
2:09 pm
himas followed right behind by the grand dragon in a dark blue suit and tie. when he entered the room and turned the corner and saw me, he just froze, and mr. kelly bumped into his back because they stumbled and regained their balance. and i knew what they were thinking, they were thinking, either they gave them the wrong room number or this was a set up, this was an ambush. so i went like this, i went to display my hands, nothing in them, but i stood up, and i approach them. i said himes or kelly, my name is darrell davis. kelly, my name. is darrell davis. >> sunday night at 8:00 eastern on q&a. >> up next, actress taraji hen son testifies about suicide risk blackntal health in youth


info Stream Only

Uploaded by TV Archive on