 And I see a question at the back. Just to re-understand this, what you're saying is the identity policy or the concept of identity is correct, but the way it is implemented is wrong. And I'm looking at it not from the UK perspective. I'm looking at it more from the US side. Look, let's nuance the idea. What I'm hopefully was trying to say was that the idea of having a ways of identifying people, and I would say particularly birth registration, is a very important feature for the inter-think. Yeah. Like in the UK perspective, in India, the obvious subsidy in India is about $50 billion. Last year, because of the fiscal year taken in March, we had about 1.5% of the GDP spent in subsidies. Since 1947 in India, about out of the British world, trade is the best in the amount of time. We have been running power-carrying programs. None of the poverty has come. Only the people who started the profiting in these programs, their poverty probably has gone according to the richest people moving around. Why is it? Because I'm serious not targeted. Today, I have a Mercedes Benz, a diesel Mercedes Benz, and I'm getting the benefit of a very subsidized fuel. Nobody wants to raise the price of the fuel. They say, oh, no, no, no, no, no, no, no, no, no, no, no. So who's getting the benefit of the fuel? So that's the reason. The only program that's for those kind of things, the larger thing, the second thing, the fuel program, that is the biggest. I think for such a thing there was a lot. The UIT program has nothing to do with the KYC, anything of that. The UIT program is just going to guarantee X Mr. Mungone Singh, who is standing in front of me, this Mr. Mungone Singh. Just three states, three states in India, my limited UIT program, have got the benefits. In India, the PTA is the public distribution system. The fraud in PTAs is about $20 billion. The fraud, out of that, 60% of the fraud because of duplication of records. So, my name, Samir Kisai, okay, I exist in 25 places. Some places I exist, like a below poverty line person, I get subsidization, correct? Now, one state, just limited to this, reduced their public distribution burden by 60%. The total cost, the total cost of the UIT program of the UIT project is not as high as in the past because what we are doing is just the UIT project. Yes, it will happen. The company is in the system. Okay, okay. My presentation is pretty rough. Okay, no, no, no, I'm happy to answer the question. My question is, so, what do you tell us in the UK experience in the world? No, I would have a question, I understand. Fine, okay. And the government is, that's fine. I'm happy to take the question. I'm happy to defer to an expert on PDS who was, I think, will want to respond in terms of the figures. Let's just think through the practicalities of the subsidized fuel. I am not aware at the moment of any intention to roll out to petrol stations, to fuel stations, a biometric reader where by every time you wish to purchase fuel, you will present your fingerprint. It will then be checked against the central database. The current, again, this workshop I was at, the current turnaround time is about 10 seconds end to end for that transaction. So you want to get on with your life. It's taking 10 seconds for the system to come back and say the fingerprint that we hopefully have matched says that you, well, actually it doesn't, it just says that you are a person. Currently it doesn't link it to your tax status. Or is it? That's stage, what you like. No, no, no. Okay, I'm going to moderate. I'm going to moderate. Okay, let's have, let's not have a one. No. Let's not have a one. No, I'm going to moderate. That's not a single thing. No, I'm going to moderate. I'm sorry. I'm sorry. I'm sorry. Sorry. I'm sorry. I'm sorry. I'm sorry. That's okay. Come ahead. Come ahead. Please come ahead. No, please come ahead. Take your mic. No, no, no. I would really like if you just ask a specific question. No, why do you have to ask that? What's wrong with you? Once again. Please allow me. Please allow me. Please allow me. Please allow me. Just ask your question. Allow me to respond. Let's not have legal reactions. Okay, my thing is, what's the main objection to the EYE project that you feel is that the way you've implemented or does that mean you should not have any or do you think of the EYE project? No. My concern is, I'll come back to the mic. The, I don't know enough about the detailed operations. I am simply based on what I have heard and have read about is I am not aware of any PDS systems at the moment using the biometric-based UID because it has to be the biometric because just giving a number is not going to do any checking is actually leading to a wholesale reduction in fraud. It might be, I'm not aware of it. I'm also not aware that the multiple recipient fraud is the greatest proportion of the fraud but this is where experts who study this area have their own more detailed evidence to provide. So that's, I don't know. I suspect that it's not as huge as you are indicating it is and I am, should we say perplexed as to how this is already stopping this from happening because as far as I'm aware it's only been the enrollment process. I'm not aware of any significant roll out of online verification services and even if that were to be the case I'm certainly not aware of any secure online verification services because if I wish to do, if you are my friend and I wish to do the fraud then I accidentally kick the power supply out when you present your fingerprint and it comes back and it doesn't come back. Oh, we assume that you're okay and we'll give you the food at that particular point. So if you want to have, you have to make sure that every single transaction is a verified transaction and that is a huge step up which will increase the costs and I'm not aware and I'm happy to be correct. One of the reasons for coming is to learn more and to, you're wrong on this and it's a differing interpretation, et cetera. I'm not aware that that is happening at the moment. So therefore I would be surprised, pleasantly surprised if it is having those kinds of effects. That's the intention, fine. My question becomes, how do you get to that intention and what are the risks all the way along the route of how that desirable intention of reducing benefit-related fraud is addressed and whether that really is the core of the problem that needs to be, that should be a focus for significant government activity. That's, I'm gonna fall back to being an academic. As an academic, those are the interesting questions and my understanding is that the evidence is not yet there, that the problem is of that nature and that the solution is having that effect. If it is, fantastic, I'll shut up. If the evidence is not there yet, that's something that needs to be studied and I look forward to hearing more about it. Actually, there just seems to be generating that from the government and it seems it's just been perpetuated by the money and it's just disappeared. What does one do? Try to move it back to a messy machine. Yep. I think this is one where I really can say, as someone in the UK who's been here for five days, I really don't have an answer to that particular one, I'm afraid. Yeah, and how you act about that. I mean, I think, again, kind of coming from my academic background, the interesting question is, okay, so you've got these intentions, these desires, there's talk about different government departments starting to include UID as one of their identifiers. Just to see how much is actually being done, when is it being proposed? What kind of coverages? What kind of benefits? What kind of costs to benefits? So separate from money being spent irresponsibly, which is technically a kind of budget oversight kind of issue. Let's assume that they are spending it in the right kinds of ways. What are the plans for the next stages of the scheme? What are the plans whereby the kinds of benefits that the previous speaker was hoping for are going to become? How are they going to be delivered? Again, when we were in discussions with the UK, there was one particular event where I was on the same panel as the lead administrator for the program. And one of the questions that I asked was, there was currently, as far as I'm aware, no plans for card readers so that you could actually use the card online to prove your identity. What is the time scale? Not how many have you got, but what is the time scale for rolling those out? Is it going to be a year? Is it going to be two years? He didn't have an answer. They hadn't even thought about that next stage because at that point, they were just interested in giving people cards. They hadn't thought about the next stage of the benefits for the individual. And that then gets a real concern from industry. Industry is saying, you're promising as this, but you haven't even given us the technical specifications for saying how we can access the documents, the verification services or whatever. You haven't told us what level of accreditation we need to achieve to be allowed to use that service directly to know whether or not you have, in the UK case, an ID card or not. So again, what are the equivalent mechanisms in place for rolling out the use of UID in these different scenarios? And if there aren't plans for those kinds of things, that's when you need to start asking the question. So we've done all of this, but we haven't even thought about the next follow-on stages. Can't we think about getting those next follow-on stages getting sorted out? This question where when you were in UK and they were, they were, so what were the inadequacies of the present identification system which led to them telling, okay, we need to be able to identify them. Okay, again, it would vary from government minister to government minister. It would vary according to their particular desires. Towards the end of the scheme, they were really concerned about young people, 18 to 21-year-olds, particularly going into bars, technically buying, in those old-fashioned days when you used to buy DVDs from the shop kind of thing, and you had to prove that you were over 18 to buy particular things. But particularly going into bars and nightclubs, there was a concern that because they've got very strict, you have to be over 18, which typically means if you look anything less than 25, they'll ask for proof of ID, which was a bit of an over-characterization. The story was that particularly women would bring their passport as a proof of ID and they would either leave them on the floor in the kind of corner of the nightclub at the end of the night because they had too much to drink and it had fallen out of their handbag, or they put it in their shoe, which was not a particularly desirable thing either, or people use their driving licence, and the problem with the driving licence in the UK is it also has the home address. So you're a young lady going to a nightclub, there's somebody on the door who's a bit sleazy but quite likes you, asks are you over 18, you show them your driving licence and that gives them your name, your date of birth, your home address, all sorts of unnecessary detail because logically the question are you over 18 has an answer that's either yes or no. It doesn't require here is my full name, my place of birth, my date of birth, my home address, et cetera. In fact, the UK ID card had the full name, place of birth, date of birth of everybody because it was also intended to be used as a travel document. It was made, it had to be printed on the front of the card and the next home secretary, so not Charles Clark, another one, when it was finally released, proudly displayed his identity card to the press and the newspapers picked up the photo of this and made a big thing saying about the British crown and the crest and whatever but it also had his full name, date of birth, place of birth. In the UK we have a good register of births and you can request birth certificates online. The only three pieces of information that you would need to give, full name, date of birth, place of birth. You can imagine what happens. Alan Arthur Johnson, made a 15th, 1952 from memory, born in London. Now if you're called Edgar Whitley and you come from a small town, it's pretty likely that there wasn't a second Edgar Whitley born in that town on that day. A second Alan Johnson, born in all of London, it's not an unreasonable possibility. So I paid my 10 pounds. I applied online for a copy of his birth certificate expecting which one of these Alan Johnson's is he. There's only one. In my office in London I have a copy of his birth certificate. Birth certificates in the UK also include a mother's maiden name. If you do online banking, typical security checks, date of birth, mother's maiden name. Not really a very smart idea to be giving all of that information available particularly for the scenario of just wanting to prove that you're over 18 so that you can get a drink. And there are very innovative technologies and I've been working closely with a company in the UK that uses biometrics. It uses biometrics, gives you a token, a card or whatever, or a chip in your phone. It stores a representation of your fingerprint on the chip. It stores the date of birth on the chip. And that's it. If you go to a bar, you put your phone on the card reader. You put your finger on the fingerprint and it goes green light, red light. Over 18 or not. It has to be my finger so I can't give you my phone and you try to enroll. I always have my mobile phone with me because I'm sad like that. But there's no database. There's no tracking. The only checking that's done is when they issue the token to make sure that my date of birth really is my date of birth. But once I've done that, they have no further knowledge of whether I'm going to the bar every single day or only occasionally or hardly ever. So in that particular case, it's stored on an NFC near field communication chip which could either be on a plastic card. So if you're travelling in London, you get an oyster card and it's on that. The newest generation of smartphones also have NFC chips on them. But obviously that's really expensive cutting edge. Or what they did as an intermediary was they just printed it onto a sticker and you stick the sticker onto the back of your phone. So it's not on the phone, it's not in the phone, it's physically on the phone or on the phone's case. But as you see, it has data birth, a code for your fingerprint, a little bit of technical detail about who issued it and that's all. Nothing about your name, nothing about your address because they don't need it. They just need to know is today's date over 18 years after your date of birth or not? Because people need, people without documentation mostly illiterate need something. So although having a central database may not really be ideal, but you feel biometric registration has a role to play, especially because when we saw and I saw some of the involvement procedures in Rural Maharashtra, people were giving wrong information that was unregistered as is. Like the same family, half of them were below power to nine, half of them were above power to nine. Some kind of recognition for that person. So far people have been using election cards, the local voting cards. Do you feel biometric registration, local registration has a role to play? Okay, so there's a couple of different things going on. So the first one is simply that all the UID does is it gives you a number. With a bit of checking possibly if you have documentation and de-duplication. For higher up in the pyramid there will be some checking. So effectively, and so then there's the question of whether the de-duplication actually works, which is both a technical question about whether the technical mechanisms for de-duplication actually work, but it also has a process question because let's say that I have enrolled more than once. Let's say that the system under two names and let's say that the system spits it out and says we've already seen this fingerprint and iris for John Smith and now they're claiming Edgar Whitley. You still need to have a process that makes sure that that application is rejected. So if I want to be a criminal and I want to get lots of UIDs for multiple UIDs for my friends, I just have to address that point of weakness. That somewhere the system will say don't issue a UID because they're already registered. Probably there will be a no actually this is unique override. If you can address that weakness in the system then I can give you three, four but they'll all be linked to your biometric. So it doesn't actually... Now I expect that UID is making sure that the people who have the ability to override and to reject aren't subject to the kinds of interference but that's a very serious risk. The second question or the second part of your question was about whether or not biometrics are needed and this is I think an area where the Indian case does reveal a very interesting feature. If you have no other documentation then a way of linking the number back to the person does potentially become an important, valuable issue. I can't go along and say look, look all over YouTube, you can find images of me that will give you confidence that it's the same person. You can't do that kind of capability if they have no or effectively no documentation. But then the next practical question comes how are you going to do that... Assuming that that's what you want how are you going to do that biometric verification? If you've had problems enrolling the person if it's a bright sunny day and you're trying to do iris biometric and you've bought an iris scanner and looked it out to the field and you've got the power supply the actual rechecking the biometric back against the person is a non-trivial task. It requires skilled staff. It requires good conditions. One of the reasons why iris biometrics have a flaky performance in modern airports is because modern airports are designed by architects and architects like light and bright open spaces. Iris recognition things want very controlled lighting arrangements. So you're going to always have practical problems so there's the de-duplication element which is a technical question and a process question to make sure it actually works and of course you've got a scale problem in India where you're talking about a billion. In the UK the inventor of iris biometric technologies so there's a certain element of bias there argued that in the UK just using fingerprints would cause lots of problems after about three or four million enrollments in that the five million fingerprint would pick up too many, it might be x, it might be y or it might be z that you'd have to have a manual process to make sure that it wasn't x, y or z. Now that clearly won't be the people who are doing crime scene fingerprints you'll have to be somebody else, more people. So he was... And he was saying that that would kick in a very low number relatively low, a few million people in the UK because fingerprints are not designed to be distinctive enough for de-duplication over that scale and lo and behold he advocated iris biometrics as being much more reliable although if you are the one billionth or the 600 millionth which is the current figure that they're aiming for person you are checking that against 599,999,999 other irises because you have to make sure that nobody previously has been done and that computationally is going to be quite significant and it gets worse the more you enroll and even then all you've done is we're pretty sure that this iris belongs to somebody within our borders officially a resident and we've given them a unique number you've still got the... and now they come along six months, six years later claiming that they have this particular number and wants to prove using their biometric that it is the same person and that's going to be horribly complicated what happens if they claim to be something and the system doesn't work or is too bright and sunny that day or there's connection problems do you err on the side of the individual or do you say no, the system hasn't said yes you are that person therefore you don't get benefits, come back tomorrow and we'll deal with you at that point the perception that the whole movement is that it also argues that there is a need for everything cards are being formed and the opportunity of person being wrong in fact one of the major plans on which biometrics in writing cards is being posed for example is that why can't you use a cheaper version of smart cards why can't you use a cheaper version of smart cards which can allow mobility to a limited extent I think I just want to build on one point which I've mentioned about biometrics the English experience does not convey the complete picture regarding the validity of biometrics in India you see the biometrics standard committee which talks about standards of the UIDA itself which is headed by the chairman of the NIC he basically said I don't know he said I don't know anything about iris because I don't have even one iota of data to depend upon to make messes iris for fingerprints he said I have no idea how this works 100 crore the maximum you have is the USAID program and that has problems but if I grow it up to 100 crore I have absolutely no idea what happened then they understand that the biometrics standard committee did a study I'll get the image the biometrics standard committee did a study of 25,000 people from the global areas of the IPHAR they said 25,000 people against 2,550,000 fingerprints they did not try this just for example 25,000 people they said that 2,5% did not have biometrics they thought it was a job another 3% they found that the biometrics would go independently up by another 3% if the software did not tune the problem even further they said women who had many on their hands simply could not be enrolled women who put many of loss of year in their mess on their hands simply could not be enrolled leave alone for 10 states now whole lot of these problems were pointed out by the biometrics standard committee then you have the proof of concept then they that's the economic tax enterprise future ULEI without any external without it's completely internal you don't know how how much those are they are giving us 93.5% accuracy after allowing for 3 tries of biometrics after allowing for 3 tries of biometrics the average maximum was 93.5% that means 6.5% after allowing for 3 tries that's the figures from the official report the first one is you have this idea of a best finger the one that's most likely to work and then the two best finger up to 3 attempts and we're still only at 99% of a billion people are able to do this now I struggle with billion people but that's 1% of a billion is an awful lot of people who cannot use their 2 best fingers on 3 attempts to prove who they are based on a small scale study 93.5% the first one that's the single best finger one is the single best finger 3 attempts 99% means 1 pro that's a huge sum of it and it's up to 3% 2 best fingers the fundamental issue here is if you are good fingers fingerprints as the fundamental identifying variable in your self-sourced what was the commonest finger one who works with chemicals agents all of the difficulties in authentic individuals I'm not looking at network breakdowns that's simply a different question that's one women with diabetes is another issue finally fingerprints keep changing okay as you grow old fingerprints keep changing particularly with manual workers if you grow older the time period within which fingerprints change come down to something like 1 or 2 years so between 1 or 2 years you have to be in good and in the meantime you may have to be in good condition and this is important because this question cannot be by say oh we have iris because in no pilot setting in no service provision is iris being used as an obstinacy everybody use the fingerprint if you are using the fingerprint as the primary obstinacy instrument of yours then you cannot run away from these kinds of things this is an internal report I take this internal report with a pinch of salt an external lot is likely to be much more much higher than the base okay and this much is for the 15 to 16 crore people that UIDN has already involved okay we don't know how much this will explode this other way it will explode when you reach 100 crore biometrics I think this project will fade under its own days and that will be called the agency of this project UID biometrics I'm pretty sure it's okay I mean and again the interesting flip side of that is if you do have those kinds of problems then potentially you don't bother with the biometric authentication which means that you're straight back to the do we know who you are give us a number that looks like a valid number yes there's somebody with that name that has that number and you've kind of okay and particularly because if you're targeting the individuals without proper documentation you have no other checks than their introducers so potentially it really unravels how do you do this in practice at the level of that's when the really detailed questions become very very important because it's that level of detail what happens if and I think the idea is that they'll tell you which are your two best fingers in case you forget please put one of these two three times still not recognising you look like you're in Edgar just go and do it let's not bother even doing it twice next time so you've got great aspirations that are undermined by the practicalities of daily operation the RBI is giving you an issue that is certain that no sorry even if the person who is your partner please go and check his address manually don't open an account without doing that because opening a bank account has a particular level of risk and they have taken a risk assessment that says the level of guarantee for self reported slash verified enrollment data is insufficient to address that particular risk which is perfectly understandable similar where we did when that project goes down and the older it's been self-contented to some local management who are the third party of their editor we go to they'll come and do the photograph simple algorithm which are made you get a card which has a female photograph for some reason so we have already had that on top we had the Span card at our section of people who are employed in the sunny government government they have this Span card so there is already a particular number of set up data authority they have this why can't we do that to an extent to build a mid-level kind of headache program and third question is why are the governments so much insisting before they have a card in hand why don't bringing the private banks and who use that platform for providing the service because the simple thing is the card is not ready why should we allow private organizations first being in government by cutting on street to the visit and most certainly most things and last thing is so otherwise I will lose track the question was why is the government planning I can't answer I am just giving you some UK experience and last question going back to the slide how telecom can leverage for example in India the mess is kind of you go to buy a new sim card the moment you get it you start using from second day onwards you start getting telecollar from that moment and then it's almost on the user and then for this way I would not like to get any calls so do you think there is an infringement happening from the first word of the word already that number same thing can be applied here yeah so the opt-in versus opt-out which is essentially the sim number issue is a really if you've read Nudge by Thaler and Sunstein is a very interesting one because any system has a default they talk about whether you put healthy food at the front if you've got a big belly you really have to reach to grab the unhealthy food whether you put people into a pension saving scheme by default and then they can choose it so thinking about default values is very very important so somebody chose whether the default value when you get a new sim is that you opt-in to receiving marketing messages or you opt-out it would appear that the decision was that you have to opt out which means that the numbers are freely available that was probably a commercial decision with lots of lobbying from the phone providers saying this is a way that we can increase our revenue stream we know that hardly anybody bothers to change their mind nobody's going to say I'm going to go onto that website and tick the box that says I want to receive it but that question about decision choice is one that really should be informing the design of any kind of thing so do you by default do this or do you by default not do that that's a very very important one and my ultimate goal is that people have at least I cared less about whether you opt-in or opt-out than whether or not you had a proper reflection on it and if it's a government policy that there's been transparency as to who's been lobbying and influencing that so I'm much more personally I'm much more interested in the decision process by which that decision was made then particularly whether it was one way or the other because I have a suspicion that if they're exposed and properly done they tend to be more in favour of the citizen back to the question of federation centralised realistically I expect but I don't know if there's any details of this the look up database that will provide the yes no answers will be more than one single database I would expect it will be state based just as a logical practicality and if so if you are here you will check this state's database first if it doesn't find you there not find you with wrong data it doesn't find you because you're visiting from somewhere else then it will ping around possibly to places where there's migration patterns or whatever so I expect that they will be federated in that sense which means that you have less data in any one particular database but then again immediately the scale of India means that you've still got millions of records even in one particular state and then there was another question in the middle your question too which I've forgotten the second question oh yes so I suspect there's probably a few different things going on at the same time if the UK is anything to go by the first one is as it were private concerns about the quality of data in different government databases so in the UK in order to get benefits and to pay tax you have to have a national insurance number there are I think the figure is something like 60 million national insurance numbers or 40 million working people now a little bit of that can be allowed for because if you are the widow of somebody their national insurance number still has to be kept live because you're still receiving the benefit that they earned but not the kind of number so the national insurance database is known to stink it's just not good enough quality data that means that other government departments that perhaps want to have relatively secure data are not think it's easier to start from scratch than to copy somebody's database and then weed out what the UK proposals are doing at the moment is actually something slightly different which is using your existence in those different databases as a way of giving confidence so if my mobile phone company wants to do an increased check on me it might check whether I have a national insurance number and it might check whether I have a voter electoral role number and it might check but it knows that it's never going to be perfect but if three good departments and one OK department all know about me they're probably happy so I suspect there's probably an element of inter-departmental politics going on why should we give our data to you so the reason it's in the UK you actually have a tax number and a national insurance number the reason why you have two different numbers is that the tax department was not prepared to pay for the lookup service from the national insurance numbers it was not prepared to go along and say here's a national insurance number is this Edgar Whitley we'll just go at our own number so you get the real politic of dirty data departmental squabbles and possibly also a this is visionary, this is forward thinking we're using the state of the art and technology we are developing and doing something that nobody else in the world is doing so I suspect there's a number of different things going on that make let's use an existing database but it's your existing database or your existing database makes it very problematic for that to happen I wouldn't copy the unless I knew that somebody else's database was really good in which case realistically this problem would never arise because everybody would just rely on the voter database or the tax database or the education database or whatever it was but I don't think you'd ever be in that situation so I wouldn't use somebody else's database I would check against everybody else's database but that only applies for the top two thirds of that pyramid it doesn't work for the people who are not documented for the people who've been given a voter ID card because it's got a female face and whatever because again this is where the scale of India and the poverty of large parts of India just become such an overwhelming factor to the whole process and to identify they usually go in terms of but there exists a partial practice of identifying people because I know I don't know because I don't know because it may have the social networks of will and it's not surprising that I mean people out there have used and developed models of that in an internet that must be in practice most of the peaks will be used but the people at large so why should the government look at those distributed identity practices as a model of identity because they are not led and they are trustworthy but along with such a large I suspect that if you are a government doing this for because it were high integrity transactions then the risks associated with that because you have no enforcement mechanisms are too high for it would be interesting but astonishing if you know someone who knows someone therefore you get a British passport to process where to work because the opportunity to game the system if you are a trusted introducer into that network such that anyone that you introduce gets a British passport the opportunities for you to be paid millions and millions of rupees to introduce non-friends becomes certainly a moral dilemma that you would face so at that level there is a very practical gaming of the system concern which I suspect is why governments are not rushing to adopt it what you do find is very interesting social practices so we were studying a website where criminals exchange details about phishing attacks and credit card details that they had obtained and botnets for attacking and etc etc there the circle of trust really works but they have really sophisticated measures for making sure that you are completely trusted and I suspect that if you turn out to be untrustworthy the penalties are going to be far more severe than a state government would be allowed to do on a wide scale so the model is very interesting the elements of that have already existed so for the British passport application you had to get the photo signed off by a clergyman, a teacher a civil servant or whatever but they don't just rely on that they also rely on these other kinds of checks so I suspect it will be part of that process and we do rely on that we do trust recommendations from people that we trust and the people we know so if Ram says this is a really good book to read so it does work but if you were to say this is someone who you should lend £10,000 to I promise they will repay from Mumbai by the way I might not be coming to London for a few more years and don't need to visit LSE at that time I might be a little bit more skeptical even though I would trust him on books to read people to meet so it's that but it does force that what is the really the level of assurance that you require what kinds of things would you trust them absolutely probably little bit wary but if everything else so it's that broader way of looking but I suspect that is that level of trust that's required or level of assurance that's required that ultimately becomes too much of a risk for governments to do at any kind of scale in fact they rely on quite a lot with the introducers we have this introducer idea the problem is that it's been gained and it's been gained so badly that it's one of the strike marks against you idea the Omen Institute just said this process is completely out of control because it shouldn't be out of control the reason why the UK cars was created in 2010 the basic is that the parties that came into power were against the idea of cars being this what do you think to what extent the arguments of the LCI and the project were considering the policy deliberation was discussed in the parliament well we know from the conversations we had with politicians of both political parties of both opposition parties we know that they were listening I can give you examples where MPs say I'm not making these statements myself I'm relying heavily on the LSE report so the political drive of the two opposition parties was certainly helped by having it's always helpful to have academic research that stands up and supports your argument I think the political fight the attacking of LSE got awareness of the issues raised to a greater level I think government losing the child benefit discs raised awareness so I was kind of expecting when the benefit discs were lost in a day or two I'd write an opinion piece saying and you do realize that the government said that they would keep your identity data secure and we warned that it would be a problem but no within three hours the press was already recognizing that there would be very practical ID cards so the president had got it if Labour had won the election and the ID card scheme had then actually continued to roll out I think then the question of the biometrics and the practicalities would have got to the situation where political upset from ordinary citizens would have increased and it would have become increasingly unpopular and people would be writing to their MP saying why am I being expected to present my fingerprints like a common criminal just to order to do X, Y or Z because fingerprinting has a very strong connotation with criminals in the UK as opposed to using it for voter registration or whatever which happens in other locations so we were expecting a longer fight with all of these kinds of concerns becoming greater such that somebody would just say let's just get rid of all of these we've got so many niggles, so many security breaches, so many people not being happy so many letters from our constituents so little buying from the banks because there's not enough people using it that is not actually etc etc that let's just quietly kill it off that's what we were expecting would have been the most likely scenario the fact that both opposition parties agreed and won and did it sped it up to a PhD student whose PhD was on the role of biometrics and he was finishing off just as the election happens so his PhD is all talking about the problems of the explanation and the examiners of his PhD said there's very little in the thesis that would expect it to have stopped for these completely independent not completely independent but from these basically straightforward politics rather as in the party supporting it got voted out of power rather than these kinds of issues that you've been warning that will be problematic over time I'm rambling, I'll stop I'll just say that you know the LSE identity project report is yet to have a strong influence on either general single reports on EYD project or other academic forms of PGM on EYD project I think it's a serious Latina I suppose in the Indian discourse on our hub that the LSE identity project report has not received the attention that it deserves in the whole discussion in that sense that's one reason why some of us decided to interview in an interview in front line in the cover I must thank Edgar for coming giving us this wonderful fascinating lecture on identity based infrastructures and from a societal point of view not just looking at it from a technical point of view but also from a societal point of view I think the lessons learned from the UK experience are central to understanding and analyzing the other project we need to get the earlier the administrators realized it the better for them because they'll probably identify the error earlier than later thank you