S2E53 AI-enhanced technology for frictionless borders

Show video

come so yeah with the liquid bomb plot I think you know there were there are multiple planes and off the order of about 2 000 people that were at risk who were who were saved by by that plot being foiled but um an example where there was a failure in a border system that could have been solved by AI is the um Boston Marathon bomb welcome to needlestack the podcast for professional online research I'm Jeff Phillips your host and I'm Aubry Byron producer and co-host today we're continuing our discussion around Ai and ocean with a discussion about the role of AI and border security this should be a great topic and joining us for the discussion is Declan Tre Zeiss vice president of global Solutions engineering at Babel Street Declan welcome to the show oh thanks for having me Jeff thanks Aubrey appreciate you taking the time with me today yeah it's been a really interesting last couple of episodes obviously our artificial intelligence is a Hot Topic these days but before we dive in to start us off um and for those that don't know can you tell us a little bit about Babel Street and your role with the company okay yeah absolutely so Babel street is a company with um with a great heritage working with government organizations agencies intelligence law enforcement and also large commercial Enterprise customers we deliver technology that allows investigators and uh people involved in um solving and helping difficult technical technological problems in the security domain in the intelligence domain and um it revolves around ocean that we're going to talk about a little bit later but also around providing very competent AI components that can deliver kind of Point Solutions and help solve very difficult problems things that we're going to talk about in in border control um but yeah understanding human language analyzing language and text and providing insights for uh for for analysts well you talked a little bit about um you mentioned border control um can you tell us a little bit about how the tool is utilized by countries in terms of border control yeah absolutely so we we actually work with some some large well-developed Western countries in helping them secure the border and we do that by providing a capability to screen and check names of people as they enter and exit the country now this sounds like it might be a relatively straightforward problem you think okay a name is a name but um I think I'll probably go into a bit more detail about how a name isn't just a name and you want the combine name with other biographic information in order to check whether it's on a watch list or a no-fly list or yeah it could even be a terrorist watch list um so yeah we solve that problem currently yeah can you backing up a minute just give us a little bit of an overview of a modern border management and what the challenges are oh yeah absolutely absolutely so I um I first got involved with thinking about borders back in 2015. uh I've been um working with a number of people across the UK home office and actually they are a customer of ours when it comes to uh the capabilities that I talked about screening names at the border um and over that time I think I've spoken to several experts in the field work very closely with them and you know my my information is second hand but I've been operating you know around around these operations and I've been delivering solutions to help um really what what is a border okay you look at the border and it's very much about what's it for well it's it's there to facilitate the the free travel of of you know innocent people that are going about their business trying to improve the economic prosperity of their their country um maybe through through trade and then through through cargo but also you know Travelers for pleasure Travelers for business those are the really you know 99.9 of all travel is is completely innocent and well intended now um your border has to facilitate that as swiftly smoothly as efficiently as possible but also it's there as a security measure it's there to prevent the Bad actors from passing over prevent the Bad actors from being in a place where they can commit crimes and atrocities that we don't want but also there are Bad actors that are bringing illicit cargo smuggling people amongst other things that are not good so your border must must be able to reject those people and identify them efficiently too so that's what a board is for now when it comes to kind of an integrated approach to border management you need to look at it from a few different angles so um okay really if you can prevent somebody crossing your border before they get there that is a as a great help and if there's information sources that allow you to get that information in advance you can prevent The Traveler leaving their destination then your border is is working effectively so pushing the Border out is um is a key feature of integrated border management using intelligence for somebody even travels um now part of the integration though is is getting data shared and visible and usable in time across many different uh agencies and and producers and consumers of data so an integrated border wouldn't just be a single isolated uh um agency or organization they'd have to work in concert with with other borders and they'd have to work with uh you know their their local defense security and intelligence agencies to um to make sure they have up to minute information and that yeah then has to get to the Border guard so they can make effective decisions so it's the integration of of people organizations agencies and then it comes down to the the problem of identity so when you have people at the border correctly identifying who they are and comparing that against the the sources of data that you have so yeah several several aspects of you know integrated border management need to be in place to be effective that's super interesting um you know for the the last several episodes of the podcast we've been talking about artificial intelligence and this podcast is about bosent um and so can we and now we're talking about border security and and I can picture in my mind a little bit of how that might All Connect but can you tell us about how Ocean or publicly available information um and the link to AI is all used in border security yeah so absolutely so I I've been around though since before it was known as ocean so in my in my early years as a technologist I I worked on um on projects for for collecting information and um but making that usable and actionable um oh since as we understand it is open source intelligence it's it's intelligence from from sources that are available either publicly or commercially and that could be anything it could be um yeah websites it could be social media uh or it could be information that you can buy from uh an organization that collates and um and organizes it but um fundamentally it's yeah it's data that's out there out there on the internet now without some degree of artificial intelligence you're not really going to be able to find something that that's key and and valuable for your um for your border to mitigate the risks in in a timely way you know you could have hundreds of analysts on every single border trying to manually scour the internet and they're not really necessarily going to find the risks that we are we're worried about so you need to have a set of analytics tools that can find the correct data but then also refine it down who the the bits that are of of Interest so identifying the people and the associated risk that goes alongside them so osin would would be a great source of huge amount of information but without the analytics tools that are often AI fueled that sit on top you're not going to be able to to see the risks that are there um and this could be things that yeah it could be analytics AI that runs across video content or something more that's more close to my heart associated with things I've been working on for for nearly 10 years now is is looking at Text data so we publish and write so much text on the internet uh whether it's social media postings on on Twitter or whether it's you know it's blogs whether it's um um instant messages between people um you cannot have a human read and understand all that so what we want to do is get to the point where we have ai and by that by which I mean language models um it could be small models they could be you know the the emergence recently things like chat TPT it could be large language models uh things that can read and understand text across multiple languages and pull out things of interest or even answer questions against that data for a in a summarized way so um that's that's where we look we look at the the convergence of ocean and and um and text analytics and natural language processing to do smart things for for yeah the border security and and as I've gone across borders and do that super fast timely as they take my name off my passport being able to do that quickly as you go through okay sifting through so the uh your your name has your name will come off the passport Jeff and um they also need to to know a little bit more about you because there's probably several Jeff Phillips out there and there's probably several ways of spelling Jeff right you spell it with a J but but somebody might spell it with a with a G so it's first of all making sure you've got the right person and then making sure that the information that you're screening against is up to date and and shows the right risks can you tell us a little bit more about that name matching technology and how it works yeah it's um it's it's something that I I talk about often so yeah stop me if I go to go too long the um it all starts with the idea that that we imagine that matching a name is a simple thing to do if the name is on the list then it's a hit but we we as humans and working with technology introduce all kinds of complexity into names all the time and uh mistakes can be made or it could even be that a name is is is perfectly correct in the way that it's written but it's just not the way that you would write your name and um complexities are added when we we take a name and we try and stuff it into a format but a machine can understand so we'll split it and break it um we we will miss Parts out you know we might lose your middle name so they become an initial or you're unfair your last name might get put first in front of your first name so by working with machines we introduce complexity and then a big layer on pop is is all about language so we are we're speaking the same language for the most part uh you know we might have a the Atlantic Ocean between us but for the most part when I say something you'll understand how to uh write that word down you'll understand what that word means but when you move between two languages there's um there's a layer of fuzziness that's added automatically because no two languages map directly and then um there can be various interpretations as to how you would write somebody's name and if you're coming from something like Korean Chinese Japanese or or Arabic for example they're they're alphabets have nothing in common with the last alphabet that we use in in Europe so um there are adaptations and standards that some people stick to but other people use different ones so names can become more complex As you move from language to language and any any system that's going to understand checking names screening names mapping them is going to have to deal with all the complexities that we introduce just from communicating names between us and then you know the ways that they can be be complicated and the phenomena of of different name variations that get introduced in those processes up was Fuzzy name matching um can you go into a little bit about that yeah fuzzy fuzzy is a slightly worrying term right we I now we find an add-on the word smart at the same time when we talk about it so you want it to be smart and fuzzy um not just fuzzy um fuzzy is how how we think about uh who names that we're checking or so a name versus a watch list or a name versus a passport yeah to verify them we want to see okay it's not just a binary answer as to yes or no does this match it's about having some fuzziness in terms of how much the the names overlap and um how confident we can be that they're they're the same or similar so there's so if if I see my name written with my first name spelled incorrectly in my middle name's missing um I can be reasonably confident the the same maybe 80 confident that those two names are the same but what we have is a degree of fuzziness that lets us say yes I think that above a certain confidence threshold I'm going to accept that these are the same rather than just uh rather than just the other binary yes or no of okay so now if if I put myself uh in from what you've seen But if I put myself in the shoes of the of of Border agents you know how do you see artificial intelligence AI power tools such as Babel Street you know how is it assisting that border agent on the ground okay so this is this is a grape so I have in my head the the idea of your the perfect border okay you know what's the perfect border look like look like the perfect board looks like completely frictionless seamless travel you you you park your car you walk in the airport you get on your plane you know you fly for five hours and then you get off and at no point where you stopped or interrupted or hindered in your in in your in your travel plans now that that would be that would be ideal but how would we how do we get there and we get there by by doing not just just one thing it's not just introducing you know an AI agent in in the middle it's it's by um by having kind of a comprehensive uh approach to it so usually when you you book a flight um you'll do it in advance you'll do it maybe a couple of weeks in advance and at that point you are you're providing information uh to the airlines you then provide it to the local law enforcement they provide it to the intelligence agencies you know all of that all of those systems currently exist um we call that advanced passenger information and and that enables some pre-screening to happen so yeah potentially you know you you could even be part of a trusted travel approach program sorry where you um you you basically give more information up front to allow more seamless travel and those programs already exist you know there are people who can travel from the UK to the US if they do that frequently they'll be on one of these trusted traveler programs um so yeah giving information in advance helps agencies screen in advance and helps you be fast track potentially if you are yeah you know traveling between countries um but there will also be times when somebody wants to travel without making plans far in advance so what would we do to get the information to help us pre-screen them we might we might look at um their behavior the things that they do on the internet we might look at the the information that they make publicly available uh about themselves out there and maybe have that fed into a system that can do some kind of intelligent targeting screening and and um fast tracking process so yeah if we if we know who somebody is maybe they you know they drive their car up to the airport they get off near the terminal maybe there is an integrated system between law enforcement a highways agency that identifies the car who the person is and says okay are they somebody that we already know is going to travel or are they somebody that we think could um you know could be about to travel and then pick in a screening process so we want to make it seamless so that somebody that approaches such an environment can be screened before they they get through the door and then we need some degree of understanding who they are when they're you know when they're walking through the airport so that the right person doesn't get pulled aside so we want to minimize the number of interventions that a um I bought a control guard would have to take so um yeah can we use a combination of Biometrics so we willingly give our biometric information up when we cross the border I'm you know traveling into the United States I um you know the first time I went there I had to give my fingerprints and um you know traveling in Asia you know I will give Iris scans or I'll give face scans that allow me to come through quicker in future so those those capabilities are getting um getting stronger all the time and obviously AI algorithms sit behind the the recognition of the Biometrics that happen so if you can identify who somebody is as they approach do pre-screening identify who they are as they walk through an airport then you're you're 90 of the way there in making this ideal border scenario and in the background you have an AI powered risk engine that assesses the level of risk associated with that person um whether it's there so on the watch list and you check their name or that you you don't have a lot less than the watch list is generated on the Fly based on risk factors that exist in the publicly available information that that's out there at the time so yeah uh almost a um but just in time risk check so you're not necessarily depending on something but that all of these things you can see require a large amount of coordination between providers of um of of the information consumers of of information law enforcement intelligence and um and the airlines as well or you know it could be the Shipping Lines you know we haven't even haven't even mentioned the idea of of um you know people traveling across land borders or sea borders carrying large amounts of goods either way I didn't even think about that as well as their two-point goods are just shipping and um what's going on there so yeah with everything you mentioned there I mean do you see AI um as being able to augment current efforts and efforts and plug into places or does does what you're talking about require a complete change in the way borders are are managed to system complete overall that's that's a good question Jeff I like I like that one so it's it's what's what's currently happening and then what might be in in the future is available well actually also in the in the past so many countries will have existing Legacy systems they'll be sitting on huge piles of data that don't actually exist in an integrated fashion um you know the worst of it might be paper tickets reports uh passenger manifests things like this that that aren't accessible or digitized so um you know and countries that have been operating as as as ports of Entry or or Transit venues they they will have a huge amount of historic information where there might be trap value so um Legacy systems even in you know that are digital won't always be connected and that can be a problem um right now the places that we see AI being implemented is really kind of Point Solutions so it's it's the algorithm that texts the name or it's the algorithm that that looks at the iris scan and makes a determination or the one that works with the fingerprint so it's a point solution for AI here and a point solution for AI there and um and although people get excited about it it's not a holistic thing it's not um the you know the the ultimate um large um large model for for borders so you know we have large large language models now where everybody's becoming very entertained by those and you know we're thinking about the the way that it might apply But ultimately once once you can move away from Legacy systems and stop thinking about Point Solutions that's where you can um you think okay really we want a risk model that that's gonna gonna be holistic and and consider all all aspects of data pertaining to to border travel Transit you know the flow of people and goods across borders and then once we have that that model developed we can ask it novel questions around the risk of individuals uh companies Etc and um and make determinations so that's that's ideally where maybe it would go um now there are developing countries that potentially can skip the step around Legacy systems so um you know if you if you've never had an integrated border system or you've never had an attempt at making a a border management system with with you know with your historic data then maybe you're in a better position to start from scratch um and not have to rely on that and I think that's quite novel so I you know I work um work with with Tony Smith he's um the um determine of of matter so the integrated border management association uh he was former director general of UK borders and and um in my dealings with him now he's he's been been absolutely um invaluable in in understanding some of the processes around the UK border when I've been working with the home office but um he often travels and holds events in different countries uh where we're at different stages in their border management Journey and and some of the more um more agile uh Nations tend to be those that are developing a new border management from a system from scratch um ultimately everybody still Raves about the fact that that the US has the best border um you know they they they they tend to do things they don't tend to do things by halves you know U.S combined you know Customs and Border Protection into a single agency and from there they they took a view that was you know integrated from the start um I could see other countries looking at going in that direction too I'd like I like the idea sorry I like the ideal scenario you talked about right yeah I think we'll do right that's what we all want yeah yeah yeah if you're a country looking I guess to set up a new system or how you can improve what kind of Technical and ethical considerations are there for taking this approach so uh I'm very much the technologist Albury so I I'm I'm not the one to give you the the ethical viewer very much but I think that we do all all worry that we're giving away too much information about ourselves all the time and it's um it's a transaction that we make so to what degree do you want to give away your your personal information for the sake of um improving your your Passage through a border well yeah I think everyone has to make like an individual decision about that but um you know there is there's a lot of information out there publicly available and maybe we're we're not necessarily aware of how much we've already given away so um you know it I I ask you know feel free to reach out to to Babel Street and some of my colleagues will maybe show you what what can be um can be found about you out there on the um in the publicly available information but yeah we we give away a lot of information and um yeah we we to a degree should should have control of how much of our personal information we give away but yeah it becomes a transaction as to how much we're prepared to give away to facilitate our own um travel that's not to say that if if somebody is unwilling to provide personal information that they are in fact a risk I think that you know never go so far as to look at it like that there's a lot of reasons why somebody might not want to give away their their personal information um so what was the the other aspect of your question I guess for countries looking to set this uh you know part of that giving away is sometimes means maybe you give away more information but you have TSA pre-check now and you get to pass it more quickly through um but yeah if if countries are looking to adopt more AI in their approach how would they how would you recommend they start going about that so um I think it's it's look at what the thought leaders in the space are are doing and um and think okay how do we we adopt that but also you know build build on top of that so um you know my my personal uh personal prioritization is is around improving screening of passengers against against watchless you know you can um you can make everything easier for everyone if if only the people that are genuine threats are pulled aside so having you know an optimal screening system that doesn't doesn't deliver false positive um alerts it's got to be high on the list this is um you're going to be able to make the most of your border guards if they're not pulling over every every 10th person um you know if every yeah it's for a single airport you could have hundreds and hundreds of false positive alerts every single day um so you're getting getting better at that you know a point solution that can that can improve your your screening uh reduce your false positive alerts and still make sure that you don't miss those those true positive alerts so it's absolutely you know real risks um is is key um so yeah I'd say go go that way first now once you have such a system that is good for screening then then you your resources are are free to to be more smart about what you do with everything else that you've got so yeah now now that you you can make sure that the people on the watch list are stopped and the people that um that aren't can can move freely well now you you think about how do I build the better watch list how do I make better use of of the information available to me so applying AI to as we described earlier you know sort through that that information from the ocean sources to um to look for new and novel uh risks that associated with it and maybe build a large model from the the historical data and the decision making that's happened um you know a supervised model plus an unsupervised model you know machine learned based off that that data um combined with um you know a decision-making engine that's gonna give your kind of future Border guard a better way of doing things because we think of Border guards as as people wearing uniforms standing in an airport or standing at a port or at a border crossing but I think that in the future that's that's more going to be um a role where somebody sits at a desk and understands you know the the general level of risk at any one time and uh can drill down into the individual risk associated with people as they as they move um so yeah assisted by a machine that can bring the the riskiest things to the top of their list yeah can you give any examples actually of AI being used to sort of tip off border agents about a potential risk I am I can I can give you um an interesting example so well I've got I've got a couple so I think probably you're all familiar with the um the limitations around taking liquids onto flights right yeah um and what we were aware there was a Netflix documentary like film documentary that that told the whole story of the liquid bomb plot um ultimately the liquid bomb plot was a plot that was foiled by intelligence agencies and um when you when you watch the the you know the film you you get to the point where they've already identified the the people that were causing um you know he's out to out to cause harm but the um a lot of the hard work happened before that so this was back in what 2006 I think something like that it was um there was already AI algorithms running against text postings on forums you know in the internet in the deep dark dark web to identify risks against um you know against individuals and who could potentially become Travelers so having AI engines that could read different languages AI engines that could understand the um the the kind of the risky terms that might be mentioned and Associate those with with people and times and flights were were um were needed to to foil that particular plot so it was natural language processing was applied on huge amounts of of collected data to pull out the you know the names of people and then the plot that they were trying to um create you know and it wasn't it wasn't simplistic things it wasn't words like bomb that they were identified those are very easy with the keyword set it some you know it's having a semantic understanding of of of individual languages and being able to um to determine you know that talk of of chemicals in in you know a different language actually those chemicals when combined together could produce a bomb so it's it's going that extra layer which um the natural language processing and text analytics can allow you to do so that was one of the kind of an early example of AI being used to troll through large amounts of data to mitigate a risk um the second example is it's not such a it doesn't have such a positive outcome so with the liquid bomb plot I think you know there were there were multiple planes then of the order of about 2 000 people that were at risk who were who were saved by by that plot being foiled but um an example where there was a failure in a border system that could have been solved by AI is the um the Boston Marathon bomb so I think that this was 2013 and um unfortunately there were two individuals um the son AF brothers who were actually already on the FBI terrorist suspect database and they still managed to enter the United States at JFK airport traveled to Boston and plant the bomb so there was a failure there in the name matching system that they were using at the time and after a senate inquiry it was understood that it was failure in the name screening capability that they had when it came to taking names written in Cyrillic so wow Russian names written in Cyrillic script and comparing them against an English watch list so that um that system after the inquiry and the identification of the the hole in the uh in the Border because ultimately it was a hole in the Border caused by a lack of capability that's when they they implemented you know after um some some testing to prove that it could be filled they implemented an AI uh smart fuzzy name matching system in fact it was ah AI based smart Aussie name matching system to secure um secure CBP after that so you know what we don't want is to have more more things like that happen that trigger a need for an AI algorithm we want to kind of preempt that if if we can but now you can know that that it won't happen again right the United States has has filled that hole uh with the best technology available um today that just made me think are like are they in linguists and your capabilities at all if you're trying to you know do these translations and match language I'm just curious yeah so um in order to build a language model that's that's appropriate for name screening or any of the other capabilities that we have around analyzing text you need to have linguists that can train the machine and they you know they need to take their knowledge and and put it into a way that the machine can understand uh build an AI model off the off the back of that so I remember at one time within the rosette team I think we counted there was 37 different languages spoken across about 80 people which was was quite impressive wow but um but also a lot of um a lot of the language capability that you know that we use the linguists that that we might use um well it you know it's about having a data team that can reach out to to agencies and say okay I need you know I need I need skilled linguists that can help me build this this model help me with this annotation task and feed that into into the machine so yeah ultimately it comes from humans you know all of the all of the learning that we'll um that will Implement into something in the technology that's fascinating yeah I think just like with that sort of Technology you automatically think okay tech people are building this but just love to hear my fellow liberal arts background people finding this place here too no it's a it's a funny intersection but you know technology and language because yeah often it is um yeah different different groups but yeah so they're quite hard sometimes quite hard to find somebody that does both but actually you can split the tasks such that the linguists can do the linguistic tasks and the computer scientists can do the the um the computer tasks awesome well Declan this has been super sorry this has been super interesting to me you know Crossing Ai and ocean and with border security um I want to thank you for your time but for before we go is there anything else you want to add about Babel Street and artificial intelligence and how uh you can assist borders to basically to run more smoothly and travelers to be less bothered so um yeah Jeff but you know I really appreciated you giving me the time to to speak today on the on the podcast yeah Jeff Aubry yeah really appreciate your time now um I think that kind of some of my my final thoughts around it is that we're not we're not done here you know it's not it's not that we have the perfect system yet we don't have frictionless orders yet and um yeah similar technology applies in other aspects of Defense security and law enforcement too they're they're not done and we're continually evolving um I think that in the old days we worried about having too much data um you know big data used to be this buzzword that we were scared of um but now we've kind of tamed Big Data um I think going going forward we're gonna have to get smarter about what we're asking of the systems and what we want them to do you know it's um it's a bit like in The Hitchhiker's Guide to the Galaxy it's like uh yeah you find out the answer to the ultimate question of life the universe and everything but what's the question that you need to be asking so um I think yeah there's a lot of a lot of thinking that that still needs to be done so we can envisage uh the best possible outcome the seamless frictionless border but um to get there yeah we can we can do better than we're doing now with the current Point Solutions um but yeah watch this space right so we're we're always trying to to innovate um you know we're always trying to bring capabilities that are novel and unique from research make them into capabilities that can be consumed as Enterprise software and you know we're not the only uh you know the only company in this space you know there's a there's a thriving Community um you know we work in partner partnership with um across um across AI across language and and complementary Technologies so if you watch this space and and look out for for yeah upcoming announcements and look out for for you know for products and capabilities that that appear and yeah we will continue to support this um this particular area yeah uh you know we've got some some very uh loyal or satisfied customers in in the Border domain and it's something that's close to our hearts so yeah so that's that's kind of what I'd pleased with well thank you again Declan for joining us today um to our listeners if you liked what you heard uh you can view transcripts and other episode info on our website authenticate.com needlestack that's authentic with the number eight.com slash needlestack and be sure to let us know your thoughts on Twitter at needlestack pod and to like And subscribe wherever you're listening today we'll be back next week with more on how analysts can use emerging Technologies we'll see you then thank you

2023-08-22

Show video