 Hello everybody, it's great to see you all here, it's really exciting to be having this webinar series on open source verification. My name is Henrietta Wilson, I research weapons regulations and I'm really pleased to have a chance to work with Olamide Samuel and Dan Plesch at SOAS on this project. I also should give a big thanks to the technical team at SOAS in particular Fadil El Abid and the SCRAP volunteers for helping make this all happen. So open source verification, we take this to mean any systems that track activities using open source data or tools and that's done by non-governmental groups or individuals. Like last time we've got another fantastic set of speakers showcasing a really diverse set of open source verification activities in terms of the sorts of organisations that are doing it, the sorts of things that they're tracking and the ways that they're tracking them. Before I introduce the speakers properly I'm going to say a few words about the overriding aims of these webinars and a brief summary of the discussion points from the first webinar. So the series has been motivated with the aim of really finding out what's going on in open source verification around the world with the understanding that technologies have transformed what can be seen around the world and who can see it. We also want to explore the possible wider significance of these isolated projects in terms of are there possible synergies between them, are there overlaps in the challenges that they could mutually face and help solve and whether it can be worth thinking about these things as components of a bigger global tracking system and if there are benefits or difficulties with doing that. The first webinar that we had I will post the details of that in the chat in a moment but it was really interesting set of discussions about specific methods, specific projects but also about wider issues. So some challenges that were picked up on in the first webinar were about disinformation and the extent to which this really muddies the task of open source verification whether it's new, the extent to which it can be overcome, the extent to which we are overcoming it right now. Commentators also highlighted issues around authenticating data that's generated from open source techniques and it's very important for the credibility of the results that they're authenticated in some way. Alongside that the projects we spoke to were very clear that they had systems in place for authenticating their own data. So there was this interesting dynamic going on there. There are also discussions about the purpose to open source verification can serve. A lot of people felt very strongly that these sorts of activities can't replace traditionally negotiated verification regimes. They're generally part of bigger treaties. In contrast to that or maybe in complement to that there was a sense that it's not intended to replace traditional verification. Rather these sorts of activities are intended to empower individuals to make the changes they need to make and support the wider verification regimes, the traditional verification regimes. So a very interesting set of conversations that maybe will carry forward through this webinar or maybe we'll think about completely new things. So what we're going to do today is I'm going to hand over shortly to four speakers. Throughout the webinar I would like you to anybody to make comments or questions via the chat function. If possible please do give me your name and affiliation and I'll be harvesting those and presenting them to the speakers when I can. The webinar is being recorded and we'll publish it on the So Us YouTube channel and it will finish at three. So our speakers today. I'm really grateful to them. I'm aware that for already very busy people a global pandemic is an especially busy time. We're going to start with Idris Ahmad from the University of Stirling and then hand over to Geoffrey Duke if he's available. Although I'm not sure if he's signed up yet but I'll be checking that. And Eric Herring who's a professor of world politics at Bristol University whose background is in international security and who now works primarily on locally led development in Somalia and will finish off with Pavel Podvig from Unidia whose background is related to nuclear weapons, nuclear proliferation and who is involved with the Russian nuclear forces. So thank you very much for hearing me out. Idris I'm going to hand over to you. Thank you. Thank you. Can you hear me? Yes. All right. So my involvement with open source started because of the conflict in Syria. I've been monitoring it from the beginning and one of the things that I noticed very early was that there was this kind of disinformation assault which I hadn't seen in the past. I've covered many conflicts in the past in Israel Palestine which very often is a source of great disinformation and also the one in Iraq. But in all of those conflicts ultimately it was possible to get to some kind of a truth but here what was happening was that in the case of Syria that there was this almost deluge of disinformation that would surround every incident that would happen. And at the beginning of the conflict it was possible for journalists to go and verify it. So the most famous first incident that happened famous in the sense that the first massacre that happened in Syria as soon as it happened it got sort of tied up in these claims and counterclaims and there was a massacre in a village called Hula and 108 people had been killed. But what happened is that pretty soon a counter narrative emerged and it was placed through different channels and the claim was made that no it was actually the rebels who carried out the massacre themselves in order to get Western sympathy or Western intervention. But that was the time when journalists were still present in Syria on the ground. So several journalists were able to one from channel four in the UK and then from Der Spiegel, Christoph Reuter, they were able to go in interview people and also look at the circumstances and verify very quickly that while the counterclaims are absolutely false. So that was cleared up soon enough. But this was around the same time when Marie Colvin was killed and the regime then started making it very difficult for any journalist to start entering. And this is the time when more and more the reliance became on people who are on the ground. They would shoot videos or somebody would take pictures and journalists became more and more dependent on Skype calls, WhatsApp and all these other manners in which they could communicate. But then the question of verifiability emerged that how do you verify all these videos or images that were coming out. So that is where it became really essential that some kind of a method emerged a verifiable method through which things could be validated that it's not just that you could say that an image is authentic, but somebody else should be able to replicate your results. And I think that that's where the OSINT community as it emerged. So it started doing this really immeasurable service. Now there were clearly cases after that in which OSINT did some really I would just mention briefly one of the case studies which was quite illustrative for me. There was obviously first was the first major chemical attack which was similarly embroiled in all kinds of claims, counterclaims, even though the facts became obvious rather quickly enough. Later on there was the one in Khan Sheikh Hun. Now this one is quite interesting to me because I was involved in monitoring this from the beginning and it was very quickly obvious what had happened because by this time in Syria there were all kinds of people who were gathering this type of information. There were people on the ground, there was something called the sentinels who were posted along hilltops who would monitor an incoming air raid and they would monitor communication. So there was that data, there was data from the from NATO which was monitoring the same airspace and so the jet had been tracked. There were people on the ground who had witnessed it. There were samples collected from the survivors. So all of this painted a certain picture which was a pretty hard picture to dismiss but very soon a big counter offensive emerged in a disinformation counter offensive. And the reason why this becomes really problematic is it's not just that it's the government behind it. It's not the Syrian regime making a counterclaim but or the Russian media making a counterclaim. They also then find amplifiers in the west and some journalists participated in that, some academics participated in that and as a result of it it appeared like there's this big controversy around it. But eventually the things that we have found right at the beginning, I at the time I had this debate with Noam Chomsky, the very influential media scholar and as a result of it I ended up writing that how Chomsky was, he at the time was amplifying these doubts and I mentioned that the reason why his sighting are quite improbable and I explained the reasons because there was all this mass of data and I wrote about it and a few months later then the OPCW validated everything that was known from the beginning because there was also a guardian journalist who had visited on the ground and verified some of the physical claims. So anyway at that time the Osint information became quite crucial. Now the need for open source had been known for quite some time, 9-11 commission after the, in its findings and later on the Iraq intelligence commission they had said at that time because one of the things that they noticed is that in the gathering of intelligence most intelligence agencies were relying on methods which were not keeping up with the available information that could be secured through much easier ways because it's out in the public. There were images that could be verified that could be used and there were other data that could be used that was out in the public sphere which could then be put together to paint a picture. But what happened is that this whole around 2010 onwards this whole subculture started emerging and it had some distinctive traits which I thought were quite which made it distinct from something like WikiLeaks because WikiLeaks had a very non-transparent model and it was not always clear what the sources were or how it was and what were the criteria through which it was being vetted and distributed and be then the 2016 election became an example in how you know it could become a source for siphoning disinformation but this was very different this culture was very much part of the whole open source movement which included things like Wikipedia and even open source software. So the thing that set it apart was its transparency that it's not just that you would find something but you would also put your methods out that here's how I got at it and you can use the same methods to test that whether you reach the same conclusions or not and I think that's what distinguished it and the first thing that they did was obviously in any of these events is verifying validating the information itself so to give you a quick example of the kind of confusions that can emerge. So when there was the attack in Khan Shekhoon and there was the chemical attack in Bhuta so what happened is that there were YouTube videos that came out and the Russian Germans soon said well this can't be correct because look at the timestamps these videos were posted a day ahead because the timestamp said you know that it was still April 3rd when the event had taken place on April 4th so clearly this is all false information except there was a obvious reason for it that all the timestamps are recorded in California where YouTube is headquartered and so because of that the date is eight hours it's actually nine hours behind so the timestamps and so Amnesty International then because of all this was able to then develop a tool it's on the website where you can always put in a YouTube video and it gives you its actual timestamp by location that whatever which part of the world it's coming from so that was one way that the information was validated and it was also some of the other features which made it quite distinct was the there was kind of a also a sense of responsibility which was crucial to your credibility within the community so to give you another example in Aleppo there was this little girl who became a focus of propaganda who was posting out videos and she and her mother would post pictures of the destruction and everything during the siege but then it became this whole campaign started discrediting the girl and saying that she was some maybe not even there and that the information that she was posting was fake and it was staged or it was from somewhere else but what happened is a true open source verification and geolocation folks from actually with some friends of mine and who are with Bellingcat and who were able to verify the girl's location and we were also able to verify which Wi-Fi networks were being used and everything so all the information was available but it wasn't released until the girl was safely out of the siege because if it was released ahead of time she the girl was actually being targeted and there had already been their home had already been destroyed so that's why it to protect her until she was safely out this information wasn't put out so there was a kind of a sense of responsibility which was also different there was a culture of sharing so there's if you go on Twitter there's some hashtag called Thursday quiz in which the open source community what it does is that it posts out a challenge either to find a location or either to do something some other task and so all these various open source um adepts they go about and whoever gets the answer they then explain that what steps they went through in order to reach their conclusion so it's a kind of a information and method sharing um uh culture which again distinguishes it and I think it's the positive aspect of it and um there's that emphasis on sharing verifiability replicability but I think that what makes it so essential for now considering especially in conflicts where access is difficult is that it answers the open source environment answers the the most enduring questions of journalism the five w's the what when where why the factual questions and I think that that can be interpreted in various ways but the thing is that the facts the integrity of the facts can only be verified with some kind of a rigorous method which the mainstream media are considering um now most of them have um created fact checking departments some publications used to have them but now there's more investment in that and but I think that it was ignored for a long time and I think that that's making a big difference that there's now greater emphasis on verifiability on uh integrity of information and I think that that has been generally a healthy contribution obviously it's not going to replace journalism but what it will do is it is supplementing journalism and sometimes challenging journalism and I think that that is what makes open source right now um to me it has been uh I see it as a mainly positive all that's not to say that it doesn't have negative uses I mean we see we have seen in the US the alt right use open source techniques to do all kinds of doxing people and doing various kinds of things which you know we would question so obviously it is something that can be misused but in terms of the mainstream manifestation human rights watch has now got an open source section Amnesty International and New York Times has got its visual investigations department I think that those are things which are positive contributions but I wouldn't say that obviously it's a tool like any other that it will always be necessarily used for the positive I think I'll end there and thank you absolutely really interesting insight into all sorts of things that you've been involved with and also some of the issues that span open source sources so I'll pick out again you're picking out this this sense of disinformation makes it much harder we had it in the last webinar but that also it enforces the open source practitioners to develop systems for authenticating their own data which are growing organically from the community and I wonder we'll come back to this in the questions and I'll hand over to the next speaker soon but I'm interested to see the extent to which is true across all open source verification and if there's something to do if there's something to be developed about code practice for people doing this sort of work yeah but thank you very much if I would just make a brief kind of addition that in terms of there is an open source community and then there's open source methods which have been adopted by right now security companies and there are all these private consultancies which are using them obviously their intention or what they intend to do with this is very different than what the mainstream kind of this community the awesome community is doing with it yes thank you Idris yeah great um I'm now going to hand over to uh Jeffrey Duke from the South Sudan Action Network on Small Arms um and I understand you've joined the meeting now Jeffrey hello Jeffrey I can't hear you hello hello nice nice to see you um over to you so tell us about your okay thanks yeah thank you thank you I have a small problem of uh network I'm not sure if it's going to remain stable but I hope that it thanks to me um thank you very much um delighted to join the conversation um I'm going to uh talk more about you know what happens at the local level in terms of open source um you know data collection verification and the use of this information in that this collection um one thing for um for start is that there is various sources of you know open source I mean uh data that is relevant to our kind of work which uh is disarmament and security um humanitarian design and uh related instruments but I will have my focus more on uh small arms or live weapons as uh something that is giving my country a big headache than nuclear weapons and all other you know conventional arms so um how do we find information about small arms and live weapons and uh what are the the legal uh you know constraints around the transfer of arms to a country like South Sudan which has been in war for the past five years or so as some of you might be aware about um some of the organizations collecting doing research every now and then and some collecting real-time information like uh you know acclade um there is a small arms survey that also provides uh research and um data sharing about arms tracking identification and responsible information sharing and then there's the eye trades um by conflict government research project funded by the EU they're all you know uh provide us with the tools for verifying information around the responsible or the irresponsible transfer of arms um in in in this region but then um the the major problem in terms of this uh data that we collect is what happens after verification uh it is it's a major challenge to the civil society community because for activists and data collectors in global south there is a real big issue as highlighted by the previous speaker of uh safety how do you verify information and share it in a fashion that does not put you at risk so I we as you may all know um the issue of what happens after data is verified is not only a regional or South Sudan specific but it's it's it's it's a global uh you know challenge in in in world politics for example even with an arms embargo in South Sudan in 2018 conflict government research established that Uganda procured um arms and ammunition from three European countries and that is in light of an existing arms embargo now when the information came out um in a conflict government research report um no one was there to take action the Uganda was at the end um the destination um um you know for the end use and um the papers were clean but this response was later on found in South Sudan but then uh what takes place after that um is what is a bit confusing uh because we don't see a standard like if somebody violates an arms embargo these are the clear things that are going to the actions that will be taken against them and it's even related to uh the the the issue of violations and uh you know it's a conventing certain provisions in the arms treaty like if a country is found to have um you know diverted arms what happens within the context of the arms treaty so that is the big question that I I would like to come with and and see how you know the open source and verification community can take the conversation forward like uh is there any way we can uh governize some some kind of consensus about uh a few steps that could be relied on after verification now um I would love to I would love to highlight a few of um you know the uh the way forward in um in in terms of advancing uh data collection verification and uh response for sharing and one is uh uh how we can pull together data source of various on various themes and uh disarmament instruments like how do we grow expertise and make it cheap and sustainable to collect data around a number of uh major themes like small arms and light weapons nuclear weapons um you know chemical weapons and and and etc uh as I've highlighted by the previous speaker also if if we do not grow the expertise and you know make it uh cheaper uh it's not going to be sustainable for the the data to continue coming up uh in the open source open source to be able to verify and use it now the second is what to trust and what not to trust this now revolves also more around who may be a trusted source but then uh at some point uh you know publishes data that causes um you know confusion um that how do how do we um in in in this space of information flowing so fast how do we make sure that at the there are sets of um instruments that we can use easily and quickly to verify data before uh action on it and how fast uh you know a country or or or a regional economy community reacts on data uh varies uh therefore there needs to be some level of uh consensus on uh how and what to trust and how do we verify uh it is kind of information with uh different degrees of confidence um I would also emphasize the safety of data collectors especially in the global south as we grow the expertise of a community of data collectors how do we also grow a global uh you know solidarity and security protocol that can keep data collectors safe especially in the global uh in the global south um in that in that in that fashion even if somebody has information uh they want to be afraid uh to share it um with sources for further verification and possibly for their action um I'm saying this out of experiences of a number of uh suicide colleagues um who shared information about uh um you know arms in intercommunal violence and not only in southern but across the region of home of Africa and sharing such information uh can you know in the end the the verification and tracing process can trace these reforms back to specific um pensions in the government and that uh you know implicates uh such people so uh they are attracting insecurity uh to uh these activists so how do we uh keep them safe I would uh great Jesse I'm going to stop there for now yeah oh thank you I'm sorry I'm going to stop there for now yeah yeah yeah um I mean it's a really yeah I want to stop there for now thank you thank you really interesting insight into real uh on the ground challenges um I think it's it's fascinating this well and uh concerning the issues you raise you raise around safety of the community of people doing the tracking um and on that note I'm going to hand over to Eric Herring uh about his project uh in Somalia uh thank you very much Eric uh thank you very much for the invitation Henrietta okay um um Somalia is not very obvious place to be talking about nuclear security um however if you speak to Somali they will uh get very exercised about the subject because um it's an established fact that in the 1980s and especially the 1990s the Italian eco-mafia um had explicit and developed plans to dump nuclear waste and uh radio and and toxic waste inland uh in Somalia and off the coast of Somalia I do say that's an established fact it's being investigated and people have been prosecuted for it in Italy there's a massive investigation in Italy over many years of this so that the fact of this is is not in doubt um what it what has never been established is whether or not any nuclear materials were actually then dumped in Somalia inland or off its coast I've done quite a lot of work on this um going through what's available and there are some kind of key candidate locations that you would look if you wanted to find this act and I worked with colleagues here including uh Tom Scott who monitors radiation everywhere from Chernobyl to Fukushima he's one of the world's leading radiation monitoring people using drones and yet Viltus who designs uh um radiation detectors but in terms of the conversation you're having today the specific aspect that that you're interested in is because Somalia many Somalis have said to me that they want they wish this could be looked into I worked spent set up this team to think about how would you look into this and indeed how could you involve the public in this process and um our team uh adapted the crayfish app which is actually really about solar radiation and so on and adapted it so that it could be used as a mobile phone radiation to symmetry app it could be used as a free download and and we tested this at CERN it definitely works and the idea is that anyone with the smartphone could download the app film a suspect object and from more than one angle it could be automatically uploaded to our database and the um the radiative or something radioactive would actually affect the digital image in very subtle ways that our software could analyze and work at the strength of the radiation and and so in that sense as an element of openness I should say it's only kind of partly open because it's not the information you give straight back to the member of the public and because you because that could be used for nefarious purposes it may be that in the end what we find is that there are no such materials available and this would be a major contribution to finally resolving this in Somali history either way but is there anything there now and it's I mean I know Somali has its many uh problems and challenges um uh but the notion but what's really struck me over my years of involvement with Somalis is this really matters to people because what we feel is that their country has been this subject of all sorts of foreign predations over the years and they would like to take control of their own environment and one of the ways of doing this would be to crowdsource this kind of reporting of suspect objects so for example if um one of the fishing vessels pulls up an object that they're worried about you could film the object using the app it once they get within the mobile phone network it would automatically upload it could dump the object back in the sea we would have a time scan stamped and geo-located and they would um they would uh the water would be good to shield them so there's a lot of ways in which this this has got advantages and in terms of where we are with this um so we've we've established this work with we've discussed it with the IEA and the federal government and so we are we're fundamentally in place with what what a system could do and it's really it's all about the process now of Somali joining the IEA and and establishing the capability and but that's that's broadly where we're coming from and the other things that are relevant here um is that Somalia has significant uranium resources not in terms of the world scale but in terms of materials that are close to the surface that again are not under any kind of regulation or control of that if Somalia wants to become a member of the IEA it has to have a nuclear security regime and that requires it to have a a means of controlling the nuclear materials within its borders currently it doesn't have those um can you share my screen for a second uh if i actually can i just i might just switch this over here add to the you're able to share screen i think i can can you see that now it's uh you mean the center if you press that so you'll be able to share your screen the green button at the bottom yeah okay where is it come so apologies um yeah you can see this now no you can't see it now you can't see the map no it's okay that's very odd okay i i'm gonna have to give up on this i have no idea why why you can't see it okay i'll just give up um if you're interested okay so i was going to show you a map of of of the relevant locations in Somalia but so i apologize for my zoom incompetence this is the first time i've ever had to share try and share something so if you want to double look at the map or look at the article in more detail um then it's in global journal called global security health science and policy um and it's uh in 2020 you can have a look at it there okay i'll hand back now thank you eric i'm now um uh just trying to get the other um i'm trying to get the link into the chat thing so uh thank you very much i'll do that shortly really interesting uh another really uh exciting sort of insight into the scope of open source techniques for meeting community challenges that maybe have not been addressed and have been ignored for many years and the sense that embedded in what you're doing is protecting the people that are doing it as well as the longer term security functions so thank you very much um i'm now going to hand over to pavel podvig um from yulida who's going to uh uh give some broader insights into open source work thank you pavel thank you very much and thanks everyone for attending i'm coming from a slightly different field uh mostly uh nuclear strategic arms control and nuclear proliferation and uh i guess every field has its own uh open source analysis and open source techniques uh and um but uh they are they are largely similar across uh across the field so and uh there are nuclear nuclear weapons and everything nuclear is usually pretty well hidden uh but nevertheless there there has been quite a bit of good work uh that has been done over the years and uh i think it is now we can we can now see and maybe throw some lessons from this field and from other uh other fields as well so uh first uh since the title of the event includes verification uh i think we should be careful and this is something that henry already mentioned uh that uh it may not be the verification that we are talking about here because uh verification is uh very clearly defined as an activity that is linked to a specific obligation under a specific treaty or an agreement and if we look at the uh the way that verification is defined uh one particularly useful definition is that uh it's a set of activities whether national or cooperative tools procedures and fundamental judgment uh about what is happening with regard uh to certain obligations in the certain agreement and uh we if we look at it from this point of view uh it is clear that open source analysis uh is definitely uh one of those things that that is uh that are used uh and in normal or uh regular uh armed control treaties open source is definitely uh is one of the tools that states use to make their judgments but it's only part of the picture and uh it is important to keep in mind that uh verification is a fundamentally political process and i would say the open source analysis is also a political process it's not just data it is uh it is the judgment that you make about this data uh is a very important part of of the picture and uh if we look at the uh experience with verification that we've accumulated over the years i think one lesson is that verification uh is most effective when it is done as a cooperative activity when people on the ground are working together they know each other they know their own attitudes biases and they build trust which is extremely important because the the the trust actually comes from data and as it also comes from the relationships that people forge and this is definitely one of the area where where is where the open source analysis or many many of its examples is probably as a weakness you can watch something from from a satellite image but you may not necessarily know what it is that people are have in mind when they do this or that and and that makes the process actually uh vulnerable to a whole kind of a political uh political manipulations if you will and not necessarily intentional in fact because for some people for example a couple of containers that are that appeared on the on the rocket test site uh is a sign of some something sinister but maybe these are just containers maybe someone brought someone is moving out or something like that so this is uh this is definitely uh there is an area uh that uh that that is uh for vulnerabilities it's not necessarily fail uh a problem for open source analysis because you can imagine ways to kind of alleviate that but by working with the local experts who know the situation on the ground but uh it's uh it's something that needs to be kept in mind uh one more aspect of uh open source uh analysis that they want to uh to talk about uh today uh is that there is a uh in the last several years maybe uh there you can see that there is a certain kind of an interesting interaction between uh open source uh can you sometimes uh yeah we're not claiming that this is uh normal practices but uh sometimes we see that uh uh open source analysis is built around uh certain intelligence data that are kind of leaked into into the public and uh in a way it is still open source and sometimes it's very good work uh no doubt about that but but in a way you see that it is uh validated uh by the fact that it is kind of uh implicitly for uh uh sort of gets a nod uh from a government and uh or from intelligent services and the problem here is uh well it is of course uh makes the process open to manipulation to a larger extent but this is something you could watch for and again there are uh ways uh to uh to make sure that the uh the analysis itself is uh is solid and I think Idris uh mentioned that there is uh there is a community that the bill around there that that idea but uh to me uh an important aspect of this kind of symbiosis is that they you know in effectively and very often it gets governments of the hook uh it kind of makes uh it possible for them to avoid kind of presenting the evidence and presenting it in in full force and sometimes uh the evidence may not be as strong as this kind of a indirect way of releasing it suggests and uh which again uh does not necessarily make the analysis wrong uh but it is uh it is uh clearly makes it vulnerable to a challenge and we've seen that uh in uh in many occasions uh that uh you you could this this kind of way creates uh could create confusion and they could be manipulated with all kind of actors uh from with all kind of intentions and this is uh this is I think an unfortunate uh part of the of the sedation uh transparency of course uh would help uh transparency with open source analysis uh as always transparency helps uh but uh it only can go uh it can only go uh that far especially when people start using uh very uh elaborate methods and tools like uh software the face recognition uh or some uh tools that detect image manipulation and things like that some of these are not particularly transparent and uh and in fact but you you can see the uh the kind of a reverse effect sometimes people when you see open source analysis uses uh this kind of a sophisticated quote for tools that that is somehow uh seen as a as a lending credibility to the analysis although maybe that's the uh that's maybe the time to kind of step back and say well maybe maybe we should be more careful about components and again uh this is not uh this is not to say that this uh this kind of a interaction is uh necessarily a failure flow but but I think this is something that uh the community and everybody who looks at the results of open source analysis should be aware of and uh we we should make sure that we are not creating a situation where all kind of actors feel you know less uh less responsible for what they did thank you thank you very much bubble what you know again well you've you've managed to sum up a whole bunch of additional things to think of um so I'm noticing we haven't had many questions or comments yet uh but from the audience so please everybody do feel free to post any questions you've got um in the meantime I'm going to be asking the questions I get to choose the questions to ask um I could I could go in all sorts of directions with all of you but I'm going to give you each one question and then go back through the speaking order um to to to give you a chance to reflect um a bit more I'm aware that time's getting a bit tight so this section will not be long in any case um Idris uh your talk spanned a huge amount a huge variety of regions and positions and political processes it seemed to me um there's one there's a set of conversations happening around so it's about the extent to which we can it's possible or desirable to think in terms of a global tracking system that's that brings together different communities and I'm going to put it to you if you think that's a sensible way to think about it or desirable or if it's better to keep it as isolated activities um Jeffrey again you know really interesting things that you gave us there you spoke about the challenges of enforcement I think and this this sense that so if you find things that are violating uh existing regulations of the arm's trade treaty or whatever what do you do then and I'd be really interested to hear more about that um in particular in the context um that I think these sorts of conversations have been happening within traditional verification regimes there's been a constant question after attribution what what do you do when you find a violation um Eric uh what a fantastic pragmatic project is going on uh to try and enable and power Somali citizens uh I'm wondering if you see it's a similar question to Idris is that scope to scaling this up joining it up to make it accessible to other people or is that not really the point of what you're doing and Pavel you you gave all sorts of interesting insights into the bigger things that that are going on um you pointed out that open source you know the one the many things that I could pick out here um is there a problem does open source verification get governments off the hook are our citizens doing things that governments should be doing which feels like a really interesting counterpoint to the sort of conversations we were having in the last webinar um about uh the fact that open source cannot replace conventional verification um and you're you're seeming to imply that maybe governments might start thinking it would replace conventional verification which has never really been uh the object so I'll stop talking and and ask Idris uh to come in uh on my very worst questions I think one of the key features of um the open source community has been reliance on crowdsourcing um if you have seen the that BBC investigation um Africa eye investigation about this um extra judicial killing of uh two women and two children in Cameroon so that investigation was able to get um made it possible to have some justice it still hasn't happened fully but some accountability and that would not have happened without some degree of crowdsourcing because it was actually the public that was able to help identify the geographical features of the location in which this incident had happened so I think that one of the strengths of open source has been the ability to rely on crowds across borders sometimes it's just a necessity of time when you're pressed for time and uh you know you're up at 4am and that's morning somewhere else so the thing is that who's available who can who can do the work so I think that there's a there's strength in that ability to be able to collaborate with people who are um across different geographies and um and sometimes that also gives you legal protection because um you know there are things that you can say within one jurisdiction but that information can certainly be used by your allies outside that jurisdiction so I think that sometimes that's also a way to bypass any kind of local restrictions so I think that yeah there are there are mainly in collaborative work and also there isn't this kind of transnational collaborative work many Eastern European projects especially working on anti-corruption work and also the sort of like the interlocking of states and various types of business mafias so they also rely on this fact that they have allies in places where journalists have more protection that who can amplify the same information so I think that yeah they're mainly I see strengths but there if there are any negative sides I'm sure there are some you know um which which can be um in the future I mean you can you can kind of um take those into account and find some kind of measures against them but mainly I think that it's a positive thing thank you that's great so there's opportunities for going global through communities interacting with each other possibly but a limitation on that because some people can't do it they don't have the protections yeah thank you yeah I could go off again but I'll I'll hand it over to Jeffrey and I was also pointing out we've got another question from a participant from an aunt thank you an aunt how can we get for I'll just put it out for anybody who wants to answer how can we get buy-in from states for this set of activities yeah Jeffrey would you like to come back with any responses about enforcement yes um okay so on enforcement uh like um you know we may all know at this point like uh a very uh uh also you know shed light on it that it is a essentially a political uh process um that's why we do not we do not see like a clear you know enforcement mechanisms into the the instruments like the arms trade treaty like the arms embargo like what happens you know when it's broken um however we can use the information that we have got for you know appealing to organizations that matter most in terms of the relevant instruments for example in the case of an EU arms embargo that has been broken by both Uganda and Sudan in the context of the South Sudanese conflict um and also seeing that the EU is pouring millions of euros in um at helping to address the humanitarian aid um we could make sense uh and in something that we have I guess with them on yeah if your arms embargo is inadequate in in terms of supporting stopping arms from reaching to South Sudan and you know that these arms are fueling conflict in South Sudan how do you still continue and remain comfortable giving millions in humanitarian aid but the rest of the bold decisions what to do next we live it to them since it's an essentially a political process but we just wish that the instruments could grow a little more too like the arms trade treaty and uh this arms embargo coming up from the security council but unfortunately not I will stop that for now right and I'm sure you're not alone that I think many people would like uh some of these things to have more teeth for that sense of internationally negotiated provisions it's very hard uh to get these things in place uh thank you Jeffrey um Eric um have you got any thoughts about scaling up um yeah so and I want to link it to Anna's great question about states being secretive about their security activities so how could be possible you know how does that work with open source verification so sometimes the state doesn't control everything that's happening and and so open source verification happens with their state rights that are not and in other sent other situations they they they can collaborate and in relation to extending the kind of activity that we're doing in Somalia it fundamentally is based on the long established relationship of trust that you build up over years of involvement with people and with successive governments so I think the idea of simply taking the kind of thing you're doing in one country and just scaling up elsewhere I think is how it isn't really how it works you've got to embed whatever technical technique you have in a set of social relations that may or may not enable that and will always require it to be adapted and uh but what we're actually we've got more potential for what we're doing is to actually extend in these sort of activities into different areas of work because once people once you have a system whereby you're gathering uh crowdsourced data on one subject and that's proven to work you can you can actually gather it in relation to significant others whether it be a drought whether it be sort of weather climate action and so on and so you can you can build out from there and that's certainly what we have been already working on great very interesting again and you know there's a really interesting theme I think emerging through all the talks today for me the sense of trust we've heard Idris talking about communities getting to know each other we talked about we heard Geoffrey talking about how important it is to build up trusted relationships with people to collect information that way uh and I think that's a really neat way of thinking about scaling up maybe not going wider geographically but going deeper using what you've got to get better and better data over more more areas yeah thank you Pavel I'm going to hand you over we uh there's just a couple of minutes left I'm afraid but it'd be really great to hear your final remarks very briefly I think it is definitely citizens should be able to hold held government accountable for their actions in that sense the open source community does a great job and it has a role I think my however open source community and citizens should not or should be aware that sometimes the governments want to to use them to pursue their own agenda and that's again this is something you can have guards or to make sure that this is not happening but that's something to to be aware of and to be to be unguarded right okay I'm sorry I misunderstood your original point there yeah the value replacing or so that's a very interesting that's maybe another setting way ways in which disinformation really complicates open source verification that people can be played uh systems uh in in the sort of data but maybe the sort of systems we've heard from Idris and Geoffrey and Eric for um uh being very careful how you build the community and how you collect the data couldn't be one way to address that I'm not saying that's easy and I'm not saying that it's equally possible in all places in the world but but maybe there are systems that can address that so it's just time now to wind up by thanking everybody all extremely for your your insights your time for the audience for being part of this Idris, Geoffrey, Eric and Pava you've given us so many interesting things to think about and I hope this will feed into the wider set of conversations we're having about open source verification just a couple of things to mention the recording of this will be on the SOAS YouTube channel the URL for that will be circulated at some time afterwards the next webinar on this theme uh in this series is it's going to be on the 9th of September and it will be advertised on the SOAS CISD events page we'll send information about that out when we can and thank you everybody and I hope you have a great rest of your day see you soon bye