New Delhi:
Nandan Nilekani speaks to NDTV's Sreenivasan Jain on the UID project, Aadhaar
Sreenivasan Jain: Hello and welcome to this NDTV exclusive. When the Unique Identity Scheme was launched, it started with immense amount of promise, partly because it was headed by Nandan Nilekani; someone who retains immense goodwill amongst all the stakeholders involved. But increasingly it is running into a storm of controversies, some external, some internal, raising serious questions about its future. Joining me is Nandan Nilekani himself to talk about the future of UID. Thank you very much indeed, Nandan.
Nandan Nilekani: Thanks
Sreenivasan Jain: Did you have any idea when you got going into this that you are going to run into such a storm?
Nandan Nilekani: Sure, obviously I did not know about the contours about what would be the situation. But when you are doing a change programme of this magnitude, which affects a billion people, we are bringing in a whole new technology and charting into uncharted territories in terms of reach and scale, there are bound to be different points of view. That's a legitimate part of the process.
Sreenivasan Jain: But you weren't familiar with the world of the government, negotiating policy and all of that. So did it come as a bit of a surprise?
Nandan Nilekani: No I don't think so. I see this as an occupational hazard of being in public life. That there are different points of view and you can't accommodate every point of view in what you are doing. But at the end of the day we process all of that and we are convinced we are doing the right thing, we keep doing what we are doing.
Sreenivasan Jain: And you are still going ahead, I mean as we speak AADHAAR, the unique identity process is kind of barreling on?
Nandan Nilekani: I think it is important to break it into 3 parts. There is one part, which is enrollment, which is getting people into the system, a one-time activity. There is a second part, which is a technology platform to ensure everyone gets a unique number. The third part is the applications on top of this which are ..
Sreenivasan Jain:.. All the benefits that are meant to accrue ..
Nandan Nilekani: NREGA payments or PDS reforms or education or you know telecom or whatever. So as far as item number two is concerned, technology is concerned, we have built a really top class platform that will process a million applications a day. A scale that has never been done anywhere before.
Sreenivasan Jain: A million?
Nandan Nilekani: A million a day, 10 lakh dedupiclations per day.
Sreenivasan Jain:Deduplication means that you can verify that a person is a duplicate or not?
Nandan Nilekani: At the time you enroll the system will check that you are, so this means that we have a 100 million people, 1 crore people in our database and 10 lakh people coming in every day. Each of those 10 lakhs has to be compared against all of the 1 crore to check. So it's a very computing intensive, very sophisticated process that has to be developed. So that's working fully and ..
Sreenivasan Jain: And you have already enrolled almost 18crores?
Nandan Nilekani: It will be close to 20 crores by February.
Sreenivasan Jain: Right and you have given away about 110-115 million IDs?
Nandan Nilekani: The technology is proven and it's working. The applications have started. We have started doing NREGA payments in Jharkhand. We are doing LPG cash transfers in Mysore; we are doing mobile KYC in Hyderabad, so on and so forth. So applications have started. The only issue is how do people come into the system, which is the enrollment part. That's where the issue is.
Sreenivasan Jain: Well there are enough friends of yours that are raising questions about the entire process. Right from the entry point, but when we are talking about the enrollment let's begin from there because that's where it all starts. And we have now been able to get a chance to look at how its working in some AADHAAR centers, is where you actually see those debates between trying to be inclusive and trying to have clean verified data, and its not always easy because people are coming up with all sort of documents. How can you verify if these are authentic or not?
Nandan Nilekani: I think we have good processes but remember we are a country where even today there are millions of people who don't have any form of identity, and we have been able to bring them into the system, otherwise they will be outsiders because they don't have an acknowledged existence. They can't apply for their PDS; they can't get a bank account.
Sreenivasan Jain: So you don't want to get trapped in that cycle where you need an identity to get an identity?
Nandan Nilekani: That will get into a loop, right, you require one document A to get document B and require document B to get document C. What about the people who have no documents? They can't enter. So we have broken that loop by creating an introducer concept and so forth. So fundamentally this project is about inclusion. It is about giving people who are left out a chance to participate, which is why we have to strike a very balanced approach.
Sreenivasan Jain: But even with this introducer system it is not an easy thing, because often we found that introducers may not know everyone in the room, so you are often introducing people on good faith.
Nandan Nilekani: But the thing is when you introduce someone we record the AADHAAR number of the introducer, so we know who introduced him. And the most important thing that you must recognize, is that this system makes sure that there are no duplicates, so a person can enroll in this system only once, and that is really the best system we can have, because if you enroll, you enroll with some other name ..
Sreenivasan Jain: So even if you enroll under a false name or with false records you are saying that the identity is fixed
Nandan Nilekani: Fixed to that name and you are going to be that name for the rest of your life. So the fact that you are establishing uniqueness, that is the real value proposition. And the uniqueness has happened because, using the biometrics that creates a unique digital signature for everyone, we have established technologically that it is a unique person. That's what is important value proposition here.
Sreenivasan Jain: So when you talk about biometrics, which is essentially your finger prints and your iris scans, which is what you record in the system, now again there, whatever little research I've been doing, suggests that this scale of biometrics has never been done and its uncharted. Even in the countries like the UK, where they tried to start an identity card system, have had serious doubts about it.
Nandan Nilekani: That's for other reasons. But anyway ..
Sreenivasan Jain:Yes it is a combination of both, both the apprehensions of an identity card, but also about the doubts about biometric data.
Nandan Nilekani: Of course they are uncharted waters. The world's largest biometric database is of 120 million. We are talking about 1.2 billion. So that's 10 times anything else in the world. We are the only project that's doing on that scale using fusion biometrics, because we combine iris and fingerprints.
Sreenivasan Jain: What's the guarantee that this biometric project of yours is going to work?
Nandan Nilekani: No, we have enough data now. We have data on 100 million people enrolled and we have done a lot of analysis, and now we are confident that our technology can scale up to a billion people with an accuracy of over 99.96%
Sreenivasan Jain:That's the figure, which you are willing to go ahead with, almost one hundred percent accurate?
Nandan Nilekani: We will publish a report shortly, which has all the data and the research behind it.
Sreenivasan Jain: And this would be an independent report?
Nandan Nilekani: No it's our report. But independently we can check it out. Now look remember one thing we are doing things no one has done but that's not a reasons for us not doing it as long as you are confident that you can get it done which we now think we are. So I don't think technology of biometric for deduplication will be an issue anymore.
Sreenivasan Jain:The other issue with biometric is, we actually got to see it in practice, where a lot of people, especially those who do manual labour, are not being able to give clean fingerprints for example, which is again a problem.
Nandan Nilekani: No, that's why we go with the iris. The whole idea of taking the iris of both the eyes was that we said the more data points you have, the more ability to ensure that there is no duplicate. The fact that we have a digital map of the iris, as well as 10 fingers, ensures that if one or two of those attributes is not the best, the other attributes makeup for it. That's why we chose this model.
Sreenivasan Jain: Nandan, the other big question I had raised about your project is on security. How secure is the data, the UID data that you have with you?
Nandan Nilekani: Very, very secure. I think in terms of the security of this system, it's really probably is as good as it can get, because right from the time of enrollment, you go and enroll in an enrollment station, the data is immediately encrypted, which means even the operator can't see the data after that. And from there, from the enrollment station, all the way to our data center, it travels completely in an encrypted manner, so nobody can hack into it. Once it comes to our system it's again out behind many security walls and the actual biometric data, once you extract, the minute it is kept offline. So there is a huge amount of things we have done to make this a very secure system.
Sreenivasan Jain: Is this data going to be shared?
Nandan Nilekani: It is a black box. The purpose of this data is only in authentication. When you go somewhere to receive a service and you say I am Sreenivasan Jain Jain, the system will reply it is him, or not him. That's all it does. Only a yes or a no.
Sreenivasan Jain: So no one else no government or private agency can demand access to this data?
Nandan Nilekani: Unless it is national security issue, which anywhere in the world you are allowed to.
Sreenivasan Jain: So for a national security issue they can access it?
Nandan Nilekani: There will be a process by which you make an application.
Sreenivasan Jain: So at the moment this data is in a box. But going down the line the idea is to converge this with other silos of information, right?
Nandan Nilekani: What will happen is that if you go to your bank and you start opening an AADHAAR based bank account, that bank account will have your AADHAAR number. If you go get an insurance policy and you use an AADHAAR number that will have. So different databases will have your AADHAAR number but they will all be different databases. That's the whole idea. This is called as a federated model so no one person has all data.
Sreenivasan Jain: With AADHAAR as the connecting link number. But at some point we also talk about the convergence of the various databases, for example the exercise going on with the National Public Register.
Nandan Nilekani: No. Everybody will have their own database. Now converge, that means somebody has to bring it together, and that first of all is not that easy to bring together databases of 2 different entities. Secondly it has to be done through a framework.
Sreenivasan Jain: So here's what the doubters say, they say that this project is coming from a development perspective. But once all the data is in the system, then it could also become a security or a surveillance project, which will violate the individuals' privacy.
Nandan Nilekani: No, the data in the system, we don't have, for example, if a bank uses it to authenticate a person's ID, and the time of withdrawal of money, the UID system does not know what he has withdrawn, how much he has withdrawn. That's only known only to the bank's system. That's what I meant by a federated architecture, data is known only to those service providers. If the PDS system uses this to authenticate a person before giving rice, wheat or kerosene, that information is in the PDS system.
Sreenivasan Jain: At the moment, you don't know, but again those who want to push this argument further, and say as long as you are talking about any government agency which controls a large amount of data on individuals, then its not entirely clear how much the government would want to know or not. And that is a problem going down the line, that we don't have enough data protection laws and legislature to allow something like this.
Nandan Nilekani: No I think, there obviously has to be a data protection law, which we have been suggesting for quite some time, not just for this, it should be there for telephone tapping, so many things are happening which are also privacy issues. Privacy is a vast subject. Also remember that privacy and convenience is always a trade-off. When you open a bank account and want to borrow some money and you want to get a very cheap loan, you'll share all details of your assets because you want them to give you a low interest rate. When you go to get a health insurance policy, you share your health details to get a health insurance policy. When you go and put your photograph on a social networking site, you are sharing details about yourself. But in each of those instances you see a certain benefit of doing that. Similarly here, when you get the AADHAAR number, the benefits that come from that are so immense, that the privacy that you give up is a very marginal thing. It's a trade-off.
Sreenivasan Jain: Okay, so you have been taking on now some of the criticism that has been coming from the outside. But you have been also facing a lot of internal opposition and a lot of internal questions within the govt. And I am referring specifically to the Home Ministry's questions that they have raised about the data that you have been collecting. They say we can't use the UID data, its not up to our standards, that's a huge problem isn't it?
Nandan Nilekani: No, I think it's like this; as I said there are three parts to this project, there is the enrolment of individuals to this system, which is a one-time activity that will cover a billion plus people; there is a technology platform that does deduplication and there are applications. Technology platforms, there is no issue, that everybody agrees that this is the right technology platform and it will be used by everybody, Applications, no issue, the only question is how do we enroll people and that's where there is a question, whether we enroll it through using the multiple register that we have or through the NPR? That's a matter, which is before the Cabinet and Cabinet will take the appropriate decision.
Sreenivasan Jain: But you agree with view of Home Ministry and the NPR that your data is somehow not up to scratch. It's not very sound data?
Nandan Nilekani: Well, we believe that we follow due process; we believe that the process is as per the process as defined by the Vitthal Committee, which we set up, which had everybody as a member. We have very high standards of security; we believe that it serves the purpose it does.
Sreenivasan Jain: The genesis of this problem is that the NPR, the National Public Register was collecting data, and UID was meant to give numbers. Why did you get into enrollment?
Nandan Nilekani: It was a question of building a model, because what is a goal? A goal is get as many as people in India, residents of India, into this system as fast as possible, so the more people offering the function the better. Today we are at 27,000 thousand locations where the enrollment is going on.
Sreenivasan Jain: But would you like to cover the whole of India with your data collection?
Nandan Nilekani: Well that is the matter that is in front of Cabinet so they will take proper decision.
Sreenivasan Jain: Some believe that you are collateral damage in a proxy war between Mr Chidambaram and Mr Pranab Mukherjee.
Nandan Nilekani: See I cannot comment on all these things. See I can tell you one thing, this project is a huge benefit for people, a hundred and seventy million people are enrolled in the system.
Sreenivasan Jain: How much is all of this costing?
Nandan Nilekani: It is costing nearly 100 to 150 rupees per person.
Sreenivasan Jain:No I am saying the overall costs involved, because you have already spent, you had a budget clearance of about 3500 to 4000 crores. You have asked for a budget extension of about 8000 crores?
Nandan Nilekani: It's actually much less than that.
Sreenivasan Jain: What's the kind of ballpark figure we are looking at today?
Nandan Nilekani: 600 million is what we are committed to before 2014, so if you want to do 600 million, it will cost you around, let's say 9000 crores.
Sreenivasan Jain: 9000 crore Rupees for the entire population?
Nandan Nilekani: No half the population. But look it at the other way. This is going to be used to improve the quality of service delivery. Now every year India spends 3000 crores every year on entitlements and subsidies, 3000 crores that's today you know, right it, will keep going up in future. And if you can bring in using AADHAAR numbers, you make sure that you eliminate ghosts and duplicate numbers among beneficiaries.
Sreenivasan Jain: So you are saying there are savings involved, which will balance out the high cost of conducting this exercise?
Nandan Nilekani: Cost is reference to its gain, so the cost benefit is very obvious.
Sreenivasan Jain: Nandan, there is also fierce opposition from this Parliamentary Standing Committee, which is meant to have looked into this bill, which is supposed to have given legislative teeth to AADHAAR. Do you need legislative teeth?
Nandan Nilekani: No, to do the executive part, to execute the project, we can do it under executive authority. We have taken an opinion from the Law Ministry and Attorney General.
Sreenivasan Jain: You do not need the Parliament's approval to go ahead with your project?
Nandan Nilekani: No, we said that from day one. We are doing that with the full legal sanction that this project, that can be done with the Executive order.
Sreenivasan Jain: But overall you feel that this is project still has legs?
Nandan Nilekani: Of course it has legs. 170 million have voted with their feet.
Sreenivasan Jain: And it helps to have Rahul Gandhi's backing?
Nandan Nilekani: Yes definitely it helps, and we think the promise of this is being seen; that this can transform services to the people.
Sreenivasan Jain: Thanks very much Nandan Nilekani.
Nandan Nilekani: Thank you