 Welcome everyone. Good evening. Thank you for joining us and thank you to my panelists for being here. We'll get to introductions in a little bit. But I thought we could start by talking about why we're here today. Essentially, a government committee has published a report on the regulation of non-personal data. And this is very landmark report. There's nothing like this anywhere in the world. So it's worth, I think, taking a bit of time to think through the implications of this report. The primary objective of this report is to realize the economic value from the use of non-personal data. We want to essentially create incentives for innovation and promote startups in India through the sharing of non-personal data. That's what the committee says in its report. And we're here today to understand what that impact will be on startups and small businesses. And we have a stellar panel to do just that. We have startup founders, VCs and policy experts, leading voices of the industry. And I'm your moderator for the session on LAN. I lead policy efforts on data and emerging tech for Google in India. Just before we get into the panel itself, I thought, you know, let's just do a quick two-minute refresher on non-personal data and the report. It's version two of the report. Version one was released in July and we had version two in December. And we're actually in the middle of a consultation process, comments are due by the end of this month. So what is non-personal data? Essentially, we're dealing with all data that is not personally identifiable. And as you can imagine, that's very broad universe. This is, you know, anonymized data at source, but also, you know, through anonymization technologies. And there's also data that does not involve humans. So you will imagine machine data would be part of this universe. So really, we're dealing with broad universe of data. So going back to why this is such a landmark report and what's discussing right now. Just very quickly to go into, you know, five or six highlights of what the committee is recommending. Essentially, they're asking for a law, a national law that would regulate non-personal data. And they're also calling for a national, for non-personal data authority to be established. So another regulator to recommend. Importantly, I think there is this idea of data sharing, which is central to this report. And there's been some progress, I think, from earlier versions of the report. Essentially, now what the committee has said is that businesses will not be required to share data with another company. But requests for data sharing can be made for public good purposes. And we'll get into that in a little bit. They also have this requirement for registration of data businesses, and we'll get into that as well. Essentially, these are businesses that deal with data above a certain threshold. And, you know, these are things like user base and percentage of revenue that comes from data. And clearly, that's a compliance requirement that, you know, at some point, all businesses might have to comply with. So we'll get into that as well. There's also this idea of high-value data sets. I think this has come from the committee on the basis of saying, look, this universe of non-personality you're dealing with is just too broad. Let's talk about what clearly the objective is and what you're trying to solve for. And they've come up with this idea of HDDs, high-value data sets, which are essentially data sets that are beneficial to the community at large. And they've given the example of things like the transport sector, the healthcare sector. And that's really what we're going to be discussing today. And I think the last thing is another kind of improvement in the report is that they're talking about now very narrow data requests, data sharing requests that can be made. I think an important point is yet, is that they've excluded proprietary data sets from the scope of data sharing requests. So, you know, no proprietary information, no inferred data sets, no derived data sets. They're all outside the scope of data sharing. So again, we're really dealing with a narrow data set essentially. But I think what we're here to discuss is with all of this said, we still need a version three. It seems like there's a lot still left to be discussed. And, you know, that's why we're here today. With that, I'm very happy to introduce the panel. Like I said, we have just experts, the leading voices from startup founders, the investor community and public policy experts that deal with these subjects from day to day. We have Deepav, who is the CEO at Lensow. We have Ashish, who heads policy at Mascom. We have Kumar, who's the CEO of Slanglabs. Prashantho, who's a public policy consultant. Sameer Varma, who's managing director at Nexus Venture Partners. And Rajan, who's a partner at Upekha. We'll also be joined by Shweta, who's head of policy at Sakaya Capital. So, very quickly, the three things that we're going to be discussing today. That we thought could be the focus of this discussion. The first one is really taking a step back and starting from first principles. What are the justifications for regulating non-personal data? From an economic perspective, an economic perspective, a social perspective. What's the model justification for this? The second theme is to deal with the trade-offs involved. An assessment of essentially what the cost and benefit is. And this is where we'll get into perhaps some of the issues around compliance. And the last is actually the regulatory modalities that the committee has recommended. And what the panel feels about what's been proposed here. In terms of the flow, we're going to get into each of these themes one by one. We'll have 15 minutes for each of these themes and we'll leave plenty of time for Q&A. We'll take questions right at the end, but please keep your questions coming. We'll take them both from Zoom and YouTube at the end. We do intend to wrap up with some key takeaways so that the committee has something to go back. So that getting right into the discussion. I'd like to start with you, Deepak, if we can, on this question of the justification for regulating non-personal data. And really, I think one of the fundamental premises of this new report is that raw data or factual data can generate economic value if shared. So where does raw data lie in this value stream? You know, as a data scientist, we're really curious to know, you know, how do you think about raw data? And what are the incentives, if any, for businesses to share this raw data with others, given the cost of collection and storage of raw data? Sure. Thanks, Andan. So before I respond specifically to this question, I do want to point out that, like you mentioned earlier, the framework in its current form at least focuses on public good rather than economic value creation as an end in itself. And maybe this is because the framework has undergone multiple revisions. And so it's come to prioritize social good now over data sharing or economic value creation. So coming to the question of raw data, I do think that data custodians or businesses don't necessarily have an incentive to analyze their data sets in the same way that someone on the outside would work good. And clearly, they obviously don't combine the raw data sets either. So just to give you an example, Strava is an application that many of us use runners and cyclists use it to track their activities. Say all cyclists are slowing down on a particular stretch of road or all cyclists are slowing down on a particular stretch of road when it rains, which is indicative of water lobbying. Now this is not necessarily something that Strava cares about or can even do anything about. But it's something that a public body might find useful and might be able to act upon. And for what is worth companies like Strava already share data with the government so it's not just an economic incentive that they that they need. But obviously the data sets that are available are less robust, not all companies are doing this and this data is only available to the government and so on and so forth. So there is potential in raw data. I don't know about value yet. And coming to the cost of making this possible. So, first of all, companies are already collecting and storing data. So I don't know if, you know, if that is necessarily what they care about. The question is really, you know, what harm can it do to a business that has gone into, you know, taken this effort of collecting this data for, you know, its own use. And any data including metadata can reveal a lot about, you know, the strategic direction in which your company is moving and so on and that can, that is I think the major concern that companies have today. But otherwise, you know, one way to think about it is that this is like your CSR obligation or a tax that large companies have to bear for the greater common good, you know, so much depends on how this is implemented and what you know how it eventually Okay, no, thank you, Deepa. Really, I think two points that you said is something I want to pick up on one is this idea of the justification seems to be social impact, you know, social good. And if I can come out, I'd like to come to you for a question on that subject. The recommendations talk about this public good purpose, and there's a lot of question about a discussion about whether that's very broad definition whether we have a good understanding of what that means. But, you know, let's take a hypothetical example, and in a second that you deal with. So assume the policy objective is increasing language diversity, something obviously relevant to India would mandating the fairing of something like, you know, NLP data sets help achieve that goal. And if not, what do you think is the right approach. Okay, thanks for this great question. So our personal belief is like, I think sharing of data is useful because then we stand out, we had the exact same problem. We were looking for such data sets to be able to now train our models to be able to know build out language, language training like there are a bunch of popular data sets available in the market, but they're very limited, especially in India, we don't get so many and then they're all hidden in some places. In notion of sharing these data sets for public good, especially for companies to be able to take them and innovate on top, definitely has value. Now the way to make that incentivize is from a business. So now we are now at a point like for example, if you take ourselves as an example, when we started out, we really wanted to be able to know collect these kind of get data from somebody. Now today we had a point where they have collected a lot of data. Now are we going to be incentivized to sharing the same data with somebody else. And last is like, oh, we're going to start suddenly start putting up our, oh, but what is the value in me sharing this data. So to be able to know, but there is definitely people who would be able to know do things more than what we have been able to do with the data that we have collected. But at the same time to be able to know incentivize us just like there's a patent system which sort of gives you a lock in time for the data that you have collected or your innovation. So if there's a lock in period of the data they are collecting, and then I maximized my initial sense. Now I want to make it available for somebody because there's no point in that sitting in my storage and not adding any more additional value to my system or to anybody else. So I think having that a delayed exposure of data, I think would be a great way for businesses to not lose initial value but also for everyone else to maximize on this data. Thank you so much, Mara. I think we are really spot on and I'd love to get some of the investors into this discussion as well because I think the point you're really talking about is where that startup is in its business journey and its growth. Deepa was making this point earlier as well. I think about insights on your strategic direction. So I'd love to get Sameer and others on this question and take on some of these first principle questions. I'm really questioning the basis for regulating on personal data. Deepa mentioned an interesting point which is that startups already share data and not to start up big businesses do this and governments do this too. So if I can just go to Ashish, Ashish has been on this committee and in part of his deliberations, I'd love to know what's the evidence of market failure that necessitates, you know, a prescriptive regulatory framework like the committee is proposing. There are already data sharing practices, you know, dense approach, a great example of that, you know, there's also examples of language data sets that are available. Is there evidence of market failure? Do we need to be doing this right now? Amlan, it's not a traditional framework where we look at, you know, a typical market failure but it's more in terms of, you know, the sum is greater than the parts. So that's really the hypothesis here. And when you put the public policy lens and located from a social good perspective, I think the potential of creating a lot of benefit for the society at large using the idea of high value, I think that's very promising. So I want to underline the fact that in all of the deliberations of the committee and even if you read the earlier version and now, especially from a startup perspective, the entire hypothesis is that this will be beneficial to the startups. So at the end of it, when we look at, you know, the report shaping up, if we don't see the benefit then I think that's something that we should worry about. And what I've heard till now also in the panel. I think what Deepa mentioned, that's fairly accurate, I think why tax could be some kind of a negative analogy, but that's what it is in some sense, you know, that you're already invested in a certain business, you're doing certain things. And if they're now, and if you take that analogy to say that, you know, reverse flowing into the ocean. So it is really the value of that ocean and what is it that we can do with that. And that's really promising. So from a policy perspective, if you also with AI and to say that, you know, what is it that we can at some point in time do to leverage this kind of a data. I think there are immense possibilities. And that's, that's really promising. So I wouldn't look at it from a narrow sense that there is a particular failure I think of where we are in the, in the entire, you know, timeline of data sharing. I think we are at very early stages. I think the examples with you mentioned or even other examples of these are very small pockets of examples that we are really deep in. So that's where I would think that there's a huge. Right, no point point point. If you want to touch on one either one other I think fundamental principle that is guiding both the discussion on MPP but primarily the data protection bill and that's this idea of consent. I would love to get you in on this question if you can talk a little bit about how you know the committee's thinking through this idea of consent as it applies to non personal data. And if you can just suppose that with society of anonymization and recommendation from the committee that users should be allowed to opt out of anonymization. What do you think this does for incentives for an engineer like to build a product where essentially you have to now create a system where users can opt out of anonymization that we hear you talk a little bit about that. So great, great point. I just like to start by stepping a little wider. And, you know, I think in terms of the question of the purpose. So public good now clearly public good is a can be a fuzzy concept it needs to be sharpened and defined more clearly. And I'd like to just say to you know the first point about possible harm to business the factors that can be harms in the public good context to and so that's just a short comment I just like to leave before jumping to the anonymization thing that we've discussed this on there was a media discussion on this where this point also came up that if you have metadata which by the way can be very much more valuable and potential for abuse etc. Then the actual date. If you have metadata on say, Stiggy or the motto and their food supplies and, you know, and you start looking at the aggregated metadata of red meat consumption in different districts. Okay now that could be for public good for health reasons, it could be misused. And it could be seriously misused if that aggregated data so it's not personal data. There's no personal issue there. That is the first part, but I think that as well as this anonymization thing you brought up is reflective of, I mean this question of did we really need this or can this is this addressable by various other regulations like you have so much regulation there, including on data sharing with the public with the regulator and contact with anonymization. I think that's a very major one and that's, I just cannot figure why this has still been retained in draft to anonymization is being discouraged here. They are saying you need to have consent for non personal data for personal data to be anonymized with an opt out being allowed. So this is like saying that okay you enter your password. Now, if you want to encrypt those password you should take consent from the user encrypt those password. The anonymization is it's a step which is taken so for example, in text when they gather personal data and then they want to analyze that for analytics for fraud frauds for whole lot of aggregated analytics, they would absolutely want to anonymize that for safety. Now suddenly we are putting a barrier here saying that no you cannot do that without consent. And so you must retain and by the way very often this kind of analytics is done by third parties and so therefore you know an entity like visa or MX or whatever would simply not want to give its personal databases to those third parties without anonymizing them. Now we are telling that that they can't. So why for analytics for a whole bunch of other reasons that you want to use this aggregated data is there and explain the additional barrier. Thank you. Thank you Prashant. Several good points that we'll come back to some of these ideas, especially the point on metadata, I do want to come back to that. But if I can just go to Samir and Rajan. So Samir is starting with you. I think that two points I think the audience will be really curious to know what the VC, the investor perspective on this is. The thing that came out from both what Ashish and Deepa said was this idea that maybe this is a tax, maybe this is a fine that data businesses have to pay to do business in India. And this wouldn't be the first example of a compliance based regulation. I'm just wondering if you share this view that at some point in time every business will be a data business. And then if that's the case, given some of these requirements around mandatory registration of data businesses, should we just accept the fact that compliance with data regulations of this sort is a cost doing business in India. So at least yeah, thanks for that. You know, the way I look at it is just talking from a capitalistic and investors standpoint of, you know, investors hate uncertainty without when they're making an investment. And when we're investing to startups, these are companies which are in the search of a business model so there are all kinds of risks thrown out there, especially in the early stages right. So what we want is, or any startup or any entrepreneur would want is absolutely good and clear regulation and, you know, less overhead when you're building out the company. So that is absolutely a given right and what what you know what we've always been as investors very office when there's a lack of clarity, you know from the government or policy or regulation. I think that can be a big disruptive force, because you know, the government could be breathing down your neck and not give you the requisite clarity when you're building out something and you know, while you think that you've achieved product market fit and you've got some bit of competitive advantage or things like that, you know that that that mode that you would have built with data and powering your business could suddenly be thrown open with some kind of regulation. So I think, as you think about this I'm glad this is in version two and you know it will probably see another version and an implementation in the next 18 months only request to regulators and all of us here is that we should emerge with a, you know, giving certainty to businesses so that it is a level playing field in some ways, but it is it is not putting anybody at a disadvantage, or the other. And the other thing is, you know, whether it's a tax, I think tax can work well when it's tangible, right. You know, if you look at telecoms they had a USO application fund, for example, it was very tangible anybody would make money put money into that into the thing and you know there's a tangible way in which things work. So the challenge here is going to be, you know, if if that tangibility framework can be put in, in some ways, that that it brings in a level playing field. And, you know, that might not be the case. And I think, you know, going back to the points on anonymization and metadata. I think those are, you know, specifically on the metadata front I think are not an expert but you know that can really disrupt the level playing field and take away any kind of competitive advantage that any company would have had. So I think those you have to take very carefully on that and most importantly, on some of these issues. The industry has to consult very closely with the government because, to be honest, the government might not be able to understand the nuances around some of these newer businesses. So you have to be very, very careful when it comes to, you know, the government also having a framework for doing this in a very objective and a very tangible and very clear way. That's what I would say at a very, very high level and reduce the uncertainty because you know you don't want to be, you don't want to be disadvantaged or putting anybody at a disadvantage but you know just making sure that things are set in, you know, in a very, very clear way. And there's a framework in which all the participants can give feedback and it's clear. Thanks Sameer, I'm hearing three things, level playing field, don't be disruptive and clarity. So I will come back to these things but I think we're also coming on the second theme that we were talking about. What are the tradeoffs? What does this mean for startups? What are the costs that they would have to incur? So Rajan would love to hear your take on all of these things that we've been talking about but also this idea of, you know, do we need to just accept the fact that compliance and some of these recommendations is the cost of doing business. So to start with quick comments on the earlier points that you touched upon in terms of themes, right? So why does regulation happen? Regulation happens for three things. Either you're trying to promote innovation or like remove the hindrance that has been created towards that. You're trying to improve consumer choices or you're trying to strengthen government or like in some cases where democracy has been weakened. And what has been seen as technology platforms have become very, very powerful. It can even sort of have sways on how democracy works. And we've seen that, you know, a few weeks ago and how Twitter was able to censor Trump, right? So that's the reason why regulation gets involved. So in that context, I would say data can be looked at as a source of value, but more than the source of value when it comes to a business, data is actually a source of more, source of defensibility. It is not as much as a river coming in or going out of ocean. It is more a fort, right? Many companies, when they become really big, they actually care about data from a perspective of, hey, does it give me defensibility so that another person does not come and like, you know, get more shares of profit. Which is what Kumar was alluding to. So data is better thought of as more rather than as a source of value. Of course, it has like raw potential value. So on your question about like the tax and the fee, I think we live in a day and age where it is much more easier for us, for anybody to be a billionaire than a millionaire entrepreneur. Or in Indian language terms, I would say it is much more easier for you to be a Mukesh Ambani than an Aamatmi founder. Because many other times when these policies are created, whether it was the NTP or whether it is, you know, related to e-commerce, right? Or whether we are talking about the bill that is in discussion. It is always used as a way to make the rich more richer. And then the smaller guy who's trying to do the innovation, he kind of gets edged out. So in that sense, right, what I would say is that, like, you know, how the bill has been framed, however the report has been framed. You need to have a special consideration for, hey, is it going to help those small businesses or the startups? Is it going to promote innovation? And there are multiple stakeholders. There is a government, there is a big incumbent, there is a citizen and there is a startup or an entrepreneur which is trying to do innovation. And that's what you'll talk about from a trade-off perspective. So for a business, you can think about it as an additional tax. I love the point that Sameer said, right? You know, you don't want to introduce another uncertainty. Another uncertainty an Ambani can deal with, but a small startup cannot deal with. So you have to make it really, really simple for him. But one of the interesting thing about data as a business is that, you know, when you think about every business becomes a data business, that means that every business becomes global business. Now, if you're going to introduce more and more complications, right? You have to keep in mind that, you know, the entrepreneurs will go to the place where there is least amount of friction. And we have seen this happen across cities where now startups are flocking away from San Francisco and moving to Miami, right? You will see a lot of countries then say, hey, here is a much better, easier framework for you to do. Of course, so then the question really to ask is when Google comes and plays in India and a startup here like Kumar is playing against Google in India, what is the rules of the game? And how is it stacked? So, so I would rather say that for India, it should be thought of as a fee to play the game. And if you're increasing the fee, then like, you know, just keep the small guy in mind. But then he's also playing a global game. And if the fee is going to be really, really high, he will move away to somewhere else where the fee is not as high. Thank you, Rajan. Really good points. I'm going to come back to a couple of the things you mentioned, especially this point about fee to play the game. And I want to get both Samir and you to weigh in on that idea. Before we get into that, I think just sticking with this idea of trade-offs, right? There was the point that Prasanthu made earlier on on metadata, which I do want to come back to in the context of trade-offs. And Kumar would be great if you can weigh in on this. So with metadata, clearly the committee believes there is again some inherent value to this. They're proposing a public directory of metadata based on certain predefined fields that everyone will, every company in this data business would have to input into. And you will talk a little bit about what you think the value of this, the utility of this could be. And as Prasanthu mentioned, clearly there are some privacy and security implications of this. So what's the best way of evaluating these trade-offs? And in this particular case of metadata, what's your view? Cool, thanks. So I think for me personally, it is like what is your starting point, right? My starting point I like to be able to look, we want to be able to know share this. I think there is value in sharing. But there is a problem in the privacy and there's a problem that we come with sharing. So we have to figure out if we do solve the problem and not start with, oh, there is a problem, let's not share. So I think the first object in my mind I think sharing makes sense of this metadata. And because a structured metadata sharing modality will definitely help. Again, like I was saying, when we started, we were scouting and scrambling for identity policy, we were going to look at structured data. It was all over the place. There was no centralized place to be able to now collect all this data. So if there is some range in which things could be available in a more structured format that we could consume, I think value creation will certainly happen. So I think there is no doubt about it. I think the challenge like a lot of other people have also mentioned is like the privacy concerns that is there, like when you want to be able to know anonymized data, for example, NLP data sets, take an example NLP data sets. Who do you anonymize? When somebody is now speaking and you want to transcribe that data and that is data set, somebody spoke and then you want to verify and then you have now collected the text that person has now spoken. Now, as a company, we try to anonymize data because what somebody could have spoken could contain personally indecurable information. Now, there are a lot of outcomes that we do to be able to know scrub out that data when we store them and when we use them for training. Now, but this is not a full proof system of when we don't try to scrub that data. But then we have additional mechanisms as a company to be able to know safeguard, making sure nobody except the machine can look at the data. Nobody has access to look at the actual raw text that somebody has actually spoken. It's only a machine that can look at, there is nowhere. Now, but we now share this data outside to somebody else. Now, who now safeguards the data that you have collected on behalf of our customer? Now, because it's now outside us, even though we think we have anonymized it, there is always a chance that it can be reanonymized because there could have been something that we had missed. So that mechanism of what is the standardization of anonymization because of the technical limitations that is, how do we now safeguard it? I think it's the concern. It can be worked out, but I think there is still a bunch of unknowns there. But if you step back and say, should we try to solve it? Should we not even try to solve it? I think we should try to solve it. We should share data. We should have the directory of everything for people to share. But we need to be able to figure out a better way to anonymize and better way to now make sure this data that we're sharing is not reversal. Thanks, Kumar. You know, it's a good panel when people disagree and I'd like that we already reached the point of disagree. Great. I think the other point I did want to make for, you know, panelists and the audiences, we do have members of the committee of experts who are listening in on this. So it would be great if, you know, as you talk about these issues, you can talk about specific recommendations you want to make, you know, we're really happy to have them be listening in on this. Swetha, I leave you here on this panel. Thank you for joining us. I'm going to bring you in on two questions. The earlier question that, you know, Sameer and Garten were commenting on this idea of, you know, compliance. And for someone who's been working in the public policy space for so long, you know, there's this idea of ease of doing business that the government really likes to talk about. So given some of what we're seeing in the data protection bill and on personal data, what do you think this does for ease of doing business in India? So that's the first question. And you can talk to some of the other things we've seen come out of that. And the second question I do want to get to in on is, I think if we take a step back, we also realize that this is a regulation of framework that also applies to the government. And I don't want to lose sight of that. So it'd be great if we can talk a little bit about what that means for both the investor community and the startup community. Do you think we should be looking at this idea of, you know, unlocking this value of data in that that's currently with the government. And I believe that there's potentially a lot of social benefits and value to be unlocked there. So should we be thinking about incentives and these justifications for data sharing differently in the context of government data? Well, all very interesting points. I'm learning and thank you so much. Apologies, firstly, everyone for being late here. But I think I want to go back and I'm not sure if we discuss this at all, but I want to go back to the entire premise of the burning non personal data because I can tell you that as as as investors and large investors in this current ecosystem, we've been having conversations with startups across, you know, startups that are well advanced unicorn startups that are starting out in our seed stage, and we've been trying to understand what this non personal data framework can mean for the startup ecosystem. And first and foremost, I think most startups are unaware of the implications of this. So I think it comes as a shocker to them when we try and help them understand that this is something that is being talked about and proposed and they find it very hard to digest. So what I can tell you is that if you're going to send out an email or a survey to startups and try and help them answer those questions. For them, the whole concept of data sharing is all about the way say a need the IO Depa is trying to do data sharing which is very different from what a non personal data governance framework is trying to put in place. So I think first and foremost, I'd like to reiterate the point that startups are absolutely unaware of the implications of what this can do. For those who have done a deeper dive and have understood this, the one answer that we get from everybody is that this is going to kill and finish innovation. Absolutely. I mean the whole competitive advantage I'm learned for any startup today is the fact that they derive value from that data and the whole idea of that collecting that data and monetizing it or doing whatever they'd like to do in terms of their algorithms or anything is the whole notion of the digital economies the whole notion of innovation that's the bedrock of innovation and suddenly you're being told you know what one finder you'll be asked to part away with that data anonymized or non anonymized personal or possible of you know those are yes we understand that we're we're we're we're trying to say that this is this is going to be data that will be anonymized it's going to be used for policy making it will be used for community business but it's very hard to tell a founder today who has spent his years of you know blood sweat, you know to try and build a startup to see you know what hand it over. So I think that concept itself is not is not something that startups are able to fathom at this point in time. You know, I think, and what's ironical is that the whole premise of the non personal data framework was based on this being a way to help the startup ecosystem. So I think we just need to go back to the drawing board and say, do we need this framework and I understand that the committee's work very hard I understand that the committee's brought about some very very important changes and some very relevant and whether it is in terms of data processors, not being obligated and all of those changes are very welcome, but I do think that we still as a as a country that is today about the third largest startup ecosystem in the world, a country that is talking about 100 Unicorns by 2025, the country that can today become the most innovative country on the planet. Now when we're looking at those kind of all big goals and ambitions, do we need to bring in this non personal data framework and upset the regulatory architecture in a way that it will spook investors is going to cause regulatory uncertainty and it's going to add to the compliance burden of startups and completely kill innovation so I just think I know it may sound strong but I still think there's a lot of rethinking we need to do around it. Now having having said that one we're moving forward to what you're saying. I think I think any amount of, you know, data sharing when we talk about has to come at a stage where we first have a personal data protection law in place and we have seen it, you know, function in its letter and spirit for a few years to understand how that is happening so we're at a stage in our data governance as well as tech policy where the country does not have a data protection law and that should be the overarching regulation that we should look at, and not only look at a data protection regulator coming in place as the as the bill suggests, putting in place rules and regulations trying to understand how that process will work. I think itself is is a huge task that that the entire tech industry the government policymakers have in front of them. In the midst of this if we come up with a separate exercise and try to do something which no other country in the world is done. I think we're just causing so much confusion in the minds of startups that are just looking for ease of doing business as you said they're just looking for regulatory certainty that just want an environment where they can thrive and flourish and have minimum regulation or at least regulation where they're expected to I mean nobody is against a privacy law everybody thinks there should be privacy so I think that that itself is something that startups are so worried about right now that that they they just they just can't not can't even imagine that these are the kind of compliance and regulatory burdens they'll have to face registering as a data business being answerable to a non personal data regulator trying to you know give out data set so I know I've given a rather long answer to your question so I think we should pause here and say that let's please you know before we get down to saying what is the best way to govern on personal data. Let's just go back and keep questioning do we really need to Go ahead. We are you loud and clear. I think very emphatic this is not good for startups right and I think the other point you mentioned is we're currently in the midst of a debate on the data protection bill that's been in the works for longer than the I think there is a point of discussion to be had on what a data protection bill that incorporates elements of non personal data would look like which is what it seems like the TPC is going down that path but maybe we'll come back to that. I do want to touch on the one big team that came out over the last 15 minutes or so which is this issue of competitive advantage and Samir and Rajan if I can get you back to discuss this question I think really interesting point of discussion here right and this touches on getting issues of trade offs but also the way in which the government is thinking about regulating it to help promote startups from your perspective as an investor. Do you see non personal data regulation as a loss of competitive advantage which seems to be clearly something that a lot of investors believe in but at the same time I also heard from folks like Kumar that perhaps it helps reduce barriers to entry in new markets. So if there is data sharing you get access to this data and perhaps you can get into a market that already has an established data. So two sides of the coin, how do you see this thing out if I can start with Rajan and Samir I'll come to you next. Yeah I think the loss of competitive advantage is a good question but the question is for whom right is it a loss of competitive advantage for incumbent or it is for the startup. I mean in this regulation discussion right and I've I've referred to things like you know the e-commerce policy or you know previously you know the telecom related policy. It was all towards the incumbent or the ones that who could pull the money to sort of set up the business right so today whoever has become really really big they treat that as a competitive advantage and they don't want to lose it. But the bill or like any other suggestion that we are looking at should be making it easy for a startup and then there is many many presidents to this not just in India but across the world. AT&T was broke up before that you know the pharma industry in the US they had a bill which again you know broke down the entire pharma industry the monopolies were reduced so most of the times when you talk about regulation it is for making sure that there are no monopolies that are getting created other than the natural monopolies that exist right. Competitive advantage for big companies data they want to protect it and they would want to sort of go forward and say hey no don't do it but but competitive advantage for startup is what one should be looking at. And if you're talking about a framework then we have to make sure that you know it is friendly towards the startup it increases competition but doesn't create like like doesn't allow an incumbent to direct more barriers. It's for whom is the question. For whom is the question yeah I think at this couple of times in your comments and Shweta alluded to this as well the committee's objectives clearly are to promote startups so let's hear the startups out in terms of what the committee is trying to do. Sameer would love to hear your views on this and then we can get to some of the founders as well for their perspective on this. So I think you know it is it is imperative that you know in today's time and age the government really understands what they're trying to get into. I think all all the big tech that has been created has been created because of capitalistic forces I agree with it. And you know that it wouldn't have been created if if the system of capitalism didn't work right and in many ways they have taken a good. They're just free and there's customer data and I agree with you know is it is it right right now. I don't know that's the case right clearly but that has created great platforms and it is it is it is where the customer data is the person is the customer in some ways and the data of that person is what is being sold right. I think the first step as you know even Shweta alluded to is that we should we need to have a strong privacy law in the country and that should be the bedrock on on which a lot of this needs to happen. I think the government is trying to you know attack this problem too many ways without really understanding a lot of a lot of the nuances under the hood right. As far as PIA is concerned or you know anonymization is concerned these are deeply hard technical problems to solve and you know I've had the good fortune unfortunately or fortunately to have studied and looked at a lot of companies in this space. I can tell you it's really really hard to do right. You can you can re identify people with a bit of metadata here and a bit of data here or there right. So I think this needs a very comprehensive. You know understanding of the problem and I think you know it's not it's not to solve it in bits and pieces right. It is that the government really needs to think about the privacy act first or privacy data law first and putting the customer or the user in control first and then you know have a framework for solving this which is not not better for it but not that should not come first in many ways because this would. You know it's a flare in some ways that you know can can can tilt the balance in the favor of of incumbents or crony capitalist. Or people who you know can enter the thing late by by not understanding it in a very comprehensive way that's what I would say that it really needs to be thought through well. I might discomfort stems from the fact that I'm not very sure whether the government really understands this in a very very comprehensive manner and you know it might set set a wrong precedence. When it comes to the major eight hundred pound gorilla in the room which is the privacy law. I think you should solve that first put the user in control and that should be the framework and that's the right framework by which you can deal with. You know the big tech and and and and who needs to be monetizing this question rather than you know a small flare of this which would set the wrong precedence and set the wrong ball rolling and you know you would have a disadvantages privacy law coming down the road. That's what I would say but so it's a complicated answer but it's a very complicated question as well. Sure right but I think there's one good message that I hear that in terms of the timing of this premature for sure we need to have a privacy law perhaps there are some learnings from there that we can apply to this so to definitely make a message. The user in control I think you can't go wrong with that I think that's the best so you're going to solve it from there and I think that's the best way to do it worldwide it's not only in India I think that will be something which will happen worldwide I think we should because we should focus on that first. So I want to come to you on a couple of questions but just before that I'd love to hear from the founders as well so Deepak we can start with you on some of the things we've heard of the last 10 minutes but especially this idea of you know loss of competitive advantage versus potentially gaining access to data and then being able to enter into market for a new product. Yeah so I think Komar made a really good point about barriers to entry right so there's certainly something to be said about making data sets available for people to use as training sets and things like that and we've seen really effective sort of product innovation come from that. As a user since the user is not represented on the panel I can say that I'm not so keen on companies owning my data just because they've given me a product that I can use either. Right so I definitely agree with the sentiment of putting the user at the center of any regulation on data, which is not what exists currently. So I think that you know there is a balance to be struck also in terms of allying the fears of new startups that they will be burdened with you know compliance and things like that. A lot of this can be addressed by setting appropriate thresholds as to whom this legislation would actually apply to and whom it would not and who can actually take on the burden of this so perhaps all startups don't need to be concerned so concerned about it as well. Got it, so thanks people so again this idea that you know users should be at the center of it and perhaps something to look at this report again from the perspective of the user. Komar your thoughts before I go to Ashish and for example. Yeah, so thanks. So thanks at the discussion I was funding my company, I'm going to disagree to work. So, I think there is this notion that like data is a new oil and so the bird has become so popular that like everyone wants to keep saying man if I lose data I'm going to lose oil. It's literally being equated to the literal oil field of the past, but just because I have oil, I can literally become a bunny because just there is nothing no no unknowns to that later. My audience is giving me money, but today data will not give you innovation. I'm going to bet if Swiggy lost all its metadata, you can't build another Swiggy. It's not possible, it's not just the fact that it's that there's a bunch of operational things that you're willing to do, there's a bunch of innovation that you have to do on top of what you do with the data. So today if you give me the entire voice data set, like when I just become the world's best voice assistant, no it's not possible. You can see a lot more that has to happen on top of that. This is the starting enabling point. It just levels some level of the enabling point. Okay, people who are interested to be able to now start this journey. Let's give them the ability to now do it because without this you can't even start the journey in the current frame of how we are operating. So you need to be able to give. Yes, there is a lot of unknown support. Like the fact that, okay, what is our organization, how do we suppose share and what threshold am I supposed to give all that is a given way. There is no doubt about that. The assumption is that it will get solved and that's where the people like us are debating and then trying to come back to the government and trying to give them better inputs to be able to solve it. But at a conceptual level, should it be even thought about or not be thought about? I think the thing is like I think it has to be thought about and I don't think it is going to defeat competitors in its raw thing. And there's a lot more to compete with just the raw data that people are providing. And I think like the Sticky's counter and so many other people have done brilliant unicorns. I think it is not just in the facts of the science that the data they're given, there's so much more to it. Okay, thanks Kumar. I think there's definitely a point there on data and what that value is. Again, to get the investor perspective on whether valuations are based on data, I think would be interesting discussion, but I know we're running out of time. I do want to get to audience questions. So Prashant, I'll come to you first and then Ashish. Prashant, you worked in the Pindex sector not a very long time. And I think the committee has a very interesting recommendation when it comes to the role of sectoral regulators. It says that sectoral regulator, let's say the RDI can develop NPD regulations that go beyond what the central NPD authority would do. What do you make of this? There's obviously the possibility for conflict and just regularly on certain things that Nusweta and Samir have been talking about. At the same time, do you believe that their expertise can be utilized here? So what's your sense of how sectoral regulators can play a role in facilitating access to HVDs? So I believe the sectoral regulator in the case of Fintech, for example, or several of them, but the RBI predominantly and then the others. They're very strong and they need to be. The National Cybersecurity Council defines Fintech, the whole BFSI space as a critical infrastructure area and the property. And so therefore the RBI has a very strong set of norms on data protection, the equivalent of privacy in that space, and sharing of data. It requires a certain flow of data on a regular basis, including a fair bit of daily data. Okay, and I think this is an example of where you're really looking for a problem or a solution looking for a problem. Because in multiple cases, now for example, in the earlier case, we were talking about in metadata, this whole case of competitive advantage, etc. And there is a strong case and which has been made, for example, in favor of the NPD authority, etc. And fourth data sharing, which is like something like Amazon, which has access to a great deal of metadata on what sells and therefore it can transfer that to an Amazon basic or something like that. But it's an antitrust and that case is going to be tackled in two places. One is the CCI and the other one is in the next few days. There's going to be a breast node, something or the other from DPIIT, which will probably knock that whole thing off of inventory. So what I'm saying is most of the problems that are being stated can actually be tackled by either sectoral regulators or CCI or DPIIT with the Ministry of Commerce. No, thanks. I think I appreciate your expertise and insights here. I think, again, this will be one of those evolving processes we'll have to see how different regulators overlap. Actually, I want to come to you for a final set of questions. One is this idea again of how we can look at what the committee has suggested in the positive light, perhaps. So there are obviously some interesting governance principles, including, for example, this idea of interoperability at making data more accessible. And I think that's definitely something that is an aspirational concept that is worth incorporating perhaps even in existing leadership practices. Love to get your views on that. And if you could talk a little bit, looking forward to what this final report could look like. And I think it seems from the discussion here that we are not ready for a final report. We definitely need a version three, but watch that version three. And then we'll get your audience questions right after that. So we have a couple of questions, but please keep your questions coming. We'll take them right after this. Yeah, over to you. And so just listening into the discussion till now, I think one of the things which we should take back as a message both to these startups and the ecosystem at large is that per se, it doesn't mean that every startup will be required to share any data whatsoever. What it is simply saying is that once a high value data set is created, and there are particular startups who could potentially contribute to that high value data set, then those startups would be asked to contribute to that data set. And it will be ensured that no proprietary business productivity link internal processes linked information is shared. So, so, so that's where it is when it comes to startup being asked to share data and they will not pay you for the data that you have collected but they will pay you for any effort that is required for that anonymization processing or sharing aspect of that data. So I think it's a very limited context where some startup might end up being asked to share a data. So that's I think the first thing which I wanted to just kind of highlight in terms of what is there in the report. I think the thing which is to respond to some years and you know about uncertainty. This is a very evolving area and very new area so the fact that there is uncertainty when you talk about non personal data is a given. I think that's, that's, that's just the way it is at present. And I think the PDP has anyway put the consumer at the center. And, and I think NPD report has in a, in a kind of a wrong way has put the consumer in the center by putting the consent for anonymization. So I think that actually puts the consumer in the center in a wrong way, but, but having made that point and what is required next. So I think the role of the data trustee needs to be carved out little more clearly, because it is the data trustee who will handle the high value data set and who will make demands on sharing of the data. So what is the eligibility of this data trustee, I think that needs to be, there is some English there in the report, and there are some principles mentioned, but I think that needs to be covered out a little bit. And I think finally what will really help is to take the discussion forward is that if the now committee can do the hard work after having done all the work that they've already put in to actually think in terms of a draft legislation. And really, you know, put that out for consultation and discussion because that will really give us a little more clarity on how this is expected to play out. Otherwise, we are looking at a report and we are trying to imagine the things as they will move forward. I think, finally, by the look of it, it is going to be a law. So if it is going to be a law then I think submitting a report and then expecting a lot to emerge out of some other, you know, quarters. That's a big jump. So I would think that just like the PDP bill, maybe invite more experts into the committee and actually think through a draft and consult on that draft. I think that will really give a lot of understanding on where we are actually headed. Yeah, no thanks, Ashish. Really appreciate it. I think just looking ahead and clearly the lots of unaddressed questions. And in fact, I'm looking at some of the questions from the audience. I already have one which is again a very fundamental question, but and you know, happy to hear the panelists respond to this, but I have one as well. So from the audience we have a question on, you know, how would the government know what data to ask for it and again this process of operationalizing implementings how is this data going to be transferred to them. And my question again would be, who decides what's propriety what's inferred what's derived data what falls outside the scope of this because again going to be potentially a conflict there and I think the incentives are this aligned so anybody from the panel would like to gain on that. So, I can have a quick go at it. So I think the first bit is clearly flowing from the public policy or public good hat. I think what the committee in the report has given us three examples, one relates to the transportation sector, and that relates to health. And that's directionally I think that's what we can look at that you know they will identify a few specific teams, and that's where they will initially start building this high value data set. And that's when the data trustee will start asking the, you know, in terms of that's where the request for data will come in. And to a second question I think it's very important point you know, this entire determination of, you know, what is appropriate or not ready. It's in two parts. So one part is where an entity probably you know refuses to share data, believing that that data is not relevant for high value data set. I think that dispute probably from my understanding of the report will go to the NPD authority. But when if I make a claim that this is propriety, I think that will take the usual, you know, process of course and which is outside of NPD to actually settle the matter there if there is actually dispute on that account, and which could be a very bad place to be in because then you might, depending on how it goes, but the other way is where it could actually do a lot of litigation, where people are splitting here on this particular aspect and I don't think we have the capacity to deal with that. It's helpful. I think definitely seems like we have a lot of thinking left to do and a lot of, I think open answers. But I think before we wrap up the two things I want to do. One is, again, we have Rahul Mathan, who has helped in the preparation of this report, who's actually come on to be able to answer some of the questions that have come up during this, I thought, very interesting and insightful discussion. And I think, you know, it's great to have Rahul here. Rahul, I don't know if you can hear me, but if you can, it would be great to have you talk for about five minutes on, you know, some of what you've heard, any responses, and then it would be great to just wrap up and have something for the community to think about. So look, this is a great conversation. Really interesting to hear. Yeah, one of the things that sort of stood out for me was, you know, the difference between the views of the investors and the startups. I thought that was really interesting because the startups are looking for data because they're like, look, I want data, you know, this barrier that Shweta spoke about. And I think maybe some of you spoke about as well. I don't know if Prashanta you spoke about that. But that's the barrier that's keeping them from doing the things they want to do. And investors actually love the barrier because they've invested in a company now they want to keep that company, you know, secure, and they want to keep the barrier. So this is the challenge. And, you know, I really like that this thing came out so well in this discussion because this is this is sort of really where the challenges and, you know, I'm not you started out saying policy is a trade off. This is sort of exactly why I want to just go back to that God analogy is enough. I want to ask God for various things at the same time someone is asking God for it to rain while at the other in exactly the same time someone saying please don't let it rain because I don't want rain right now. And so in the policy making process, all the time someone's going to be unhappy. But I think the, the, the reason why these sorts of sessions are good is because we can listen to all of these sort of these different aspects. And then, you know, I am very recently on this side of the panel. I in version one actually was on that side and commenting on the report. So it's a bit, it's a bit weird for me but but look I think all all that, you know, I'd like to say is that we certainly do listen to all of this and these inputs are actually very valuable. A couple of responses to some of the questions look I think the way in which the metadata is going to work. I think the, the, I just realized I'm not on video. Okay, there you are, I was talking away trying to figure out. So the way in which, you know, companies are going to access data is essentially through that metadata registry. The idea is look if you're to create a high value data set, when you start. And again, if there are better ideas I think certainly it will be really useful to listen to those ideas. But if you want to start create a high value data set. And if we're saying that that is a public good and that we should encourage the creation of that. How do we know what goes into that data set. One way to do that is to say, look, all you companies collect data, please list the types of data that you collect and you know, don't list exactly what it is but say look we collect this sort of data. Then out of that there would be some person who's thinking completely out of me. I think the example that Deepa probably gave of Strava is is great. And you know you could potentially get that sort of data set through not just Strava but maybe Strava plus Ola plus Uber plus, you know the local transport company whoever's got GPS from their vehicles and and then come up with a data set that says where is their water logging in the city. Now that is a public good and but how do I know that I can get this public good if I don't know that Strava collects this information if I don't know that the bus transport company collects this information. And the idea was to keep to keep a registry that you can reference and say okay this is the data that's available and this is how we're going to do it. And obviously we don't want to disincentivize companies from creating data sets because if we do that then I would really be killing the goose that lays the golden egg and this is once again the balance that has to be that has to be arrived at. So if there are if there are better answers to how to do this absolutely please do send it in. I think, as you noted, the focus of this report has been narrowed to say specifically that it is focusing on public good. I think that was an ambiguity in the previous version and I and I hope that that has by narrowing the scope has given more clarity and comfort to people. I actually don't like the tax analogy because you know I hate paying taxes. I pay all my taxes but when I do it it really hurts. So if there's a better analogy for this more along the lines of, you know, we're sharing I like the reverse to the oceans kind of analogy, because we all benefit from the reverse we all benefit from the oceans. Ideally that is that is what has to be done and we can't dam the rivers we can't, you know, create a little private lake that we can all use the waterfall that's sort of the, the type of analogy. So I'm going to stop there I know as I know is very particular with time and I'm one minute short of your end of time, but but really great glad to have listened to all of this and thank you all for your inputs. I would really appreciate that and essentially you've done most of my job which was to bring this out together have some takeaways, we would have shared this with you and rest of the committee but I think you've done a good job of that I think just to tie up some news ends. I think this point very early on, right, which is that now public good seems to be the fulcrum around which all of this is going to happen but that could result in harm. And I don't know if we've thought through them with issues like metadata but you know, I think there are lots of potential use cases, definitely something to think about this value of data I think you know very early on there was some discussion but there's an economic lens you can use to that this is social lens and now, you know, clearly we see even within this panel, investors look at value data differently startups look at it and you know clearly I think the government and big companies could look at this very differently so worth looking at it from that perspective as well. I think should have made a really good point which is that startups even aware of this, the implications of this, how do we make them more aware of what this could mean for them. And I know some of the VCs the investors in this room will go back and say hey there's something coming you should be watching out for but I think you know folks like she is from NASCAR and perhaps the government needs to, I think, create a little more awareness around this. And this issue around regularly if you certainly came up and I think the recommendation that was very clear. We want certainty right now there's this discussion on the data protection bill, we need to have that first, you know, past learn from that before we can even think about any regulation and on personal data. There's still some fundamental, I think, concepts around even things like you know data ownership that came up which I don't think we have good answers around so definitely I think warrants more discussion and thinking. Really interesting point was around users and where is the user in all of this that we started from this assumption that all of what the committee is suggesting is good for startups, and it seems like not entirely the case. And perhaps if we get enough, you know, user groups in this room we might find that issues that we're proposing a recommendation to make an anonymization aren't actually good for you. So I think there's definitely something the committee should think about. And lastly, I just think this idea that there is so many open ends on you know how this is going to be implemented and operationalized. That definitely I think the point was we need to see a version of that get into more detail. So that's it I really want to thank everyone for staying around for an hour plus on a Friday evening. Really appreciate it as I said stellar panel. I learned a lot from this and a lot of our audience was in comments already. I really appreciate this conversation and we are also wrap this up. Thank you again.