J1 RF 09 Panel Names & Identity

all right welcome to this afternoon's panel I am again and still Nick Johnson for the etherium Foundation and the ENS aetherium name service project I'm just going to get the other panelists to give brief reintroduction and then we'll get on to questions hello again undressed you may to you port well I'm still philosopher in tech working on a term decentralized digital identity through services and/or transferring tech working and edits all right I'm gonna start with a couple of really easy questions for the whole panel and then all open up to you guys in the audience just if you want to ask question raise your hand and we'll rush to you with a microphone so my first question for all the panelists is what do you think the biggest barrier is to getting widespread adoption of decentralized identity today well I think we're following this through mass adoption we're facing the same problems it's hard hard technology for people to understand right now but as as they're seeing the value of it I think adoption will be coming together so to me is the availability of of a very very simple usable wallet because today as I said before beta mask is working also kind of wallets are working even ledger is working but still very difficult to use and to configure and it's out to for people to understand that if they lose their keys they can lose all their money so we really need a solution that handle all these problems you know use of handy way to make blockchain accessible to everybody this is a biggest bio I think yes for the use of perspective I think as a wallet and the way to all the key and to securely keep the key is a big problem and for the developer perspective the cost of developing the brain some smart contact and there's a collection custom an interim can also be problematic for her for some use cases second question is given the fact that the identity solutions that most of you are pursuing involve trusted Authority who's signing a a certificate for a particular identity how important is it to focus on reducing trust in the rest of the system is that still necessary and what links do you go to in order to do it and I'll start with you this time the objective of the platform like edits is to you don't have to twist one entity or another organization you have to choose who you can trust and who you cannot trust and it's a free to the user free to the service provider to accept some entities and to accept to to have trust in insolence it is an organization yes so the trust the trust is the trust on information is between two actors the one who own the information and the one I'll give you some information and I want this information to be accepted to be trusted by this guy so it depends after I think two things the legal value of the info of the of what I'm doing is still important because if if I want to have legal value behind the transaction as I Mickey I'm making I must have a trust a short party that is recognized by the court if if I have problem if I ever if I I'm in front of the court I must be able to prove that the information I give us the right one and then in this case today at least I have two to twelve somebody that is trusted by the court itself before so it's a legal entity - it could be located notary etc etc otherwise if I want to if I ever I mean an exchange with anybody and I want him to trust me we can use what is the same kind of decentralized test that teaspoon that is that has been created by you pot I don't know if it's by you pot or by somebody else but the first time I saw that is with you pot where are you you give the ability to few people's to give trust on your information if these people independent enough of you and of the guy you want you are giving your function you can do that so do they today you have different kind different way of of giving these trusts you go the goal is to use only blockchain and no other platform in the trust in the trespass and our story depends on do you or do you want to have legal value or not to to to know who are so parties you have to trust yeah I agree with them basically what we're doing here is like we are decentralized the trust on an identity right now we only need to we trust our governments or trust different corporate a tester for bringing the identity information but with these new technologies we can have a group of friends attest something for example if instead of reading of a good what what is a good restaurant here in Paris preferred green attestations from some friends instead of very specific magazine or something like that so that will allow a completely different way of interacting with identities and trust between us and also some right now we trust this source but in the future probably we can trust some sources but also we can add trust to by quantity for example the refugees sometimes data they came and they don't have any idea or anybody that can attest their names or the birthday but they come with the other hundred refugees so for a NGO probably it's more important that the they have hundred attestation saying this is that their name instead of saying this government attest that's the name so we can build models based on the number of other stations instead of the source of the attestation and that so that will change the game completely for for many use cases all right questions from the audience if you have a particular person you want to address that say otherwise we'll just open it up to the panel yes so a question for the whole panel with off chain private information which is verified on chain there is a problem when the company that has received the private information leaks by accident or on purpose the information because then it could be made public forever how do you think we can solve this problem I don't think we can solve that problem if I give you my name if I give it encrypt or whatever you should decrypt it somehow there's no way to prevent there's no way I can prevent that you will store their information or link it somewhere else we are building technology that helps you not to store that but there's nothing that prevents you to store that also sometimes you don't need to provide the whole information for example if you want to know that I am older than 18 or 21 instead of I giving you my birth date maybe you can query in a memorial encryption way how old am I without knowing exactly my my birthday but as long as you know I'm old you can still lick that information so sorry we don't waiter we cannot prevent that I mean I don't know how to maybe I agree that it's not there is no way to solve it definitely but one one one passenger in this direction is instead of asking for the information you want to check you ask for the check itself directly from Japan instead of asking for the age of the guy because you want to access some certain specific website you ask him are you an adult or not so you don't know the age if I want to open a bank account I don't need your name your I just want to know if you are allowed to open a bank account so if you are not a terrorist if you are not politically explicit people etc etc if you are not in bankrupt but I don't know the name I know the profile and maybe it's to go from a list of very small granularity information to certain kind of profile and this kind of profile we will be anonymous but use only thing you know is that this guy behind this ROM address is able to open a bank account is this list of countries you don't know who is he you don't know if his adult or not you you just know that you can do that or not and for lots of from lots of kind of of claim you you need it can be sufficient without disclosing all the identity of everybody hi so I would like to turn this in a little bit different way that in my opinion like that the centralized identity in the decentralized and there is always a trade-off between privacy and the fallback is we can't have both like it's either fully private and decentralized or we provide some fallbacks hence like this solutions we've described here not work for a mass adoption so what do you think about like adopting at different kinds of identity like using biometrics something that will identity myself not in an artificial way but something that myself I'm a fallback I mean others may be able to answer this better than I but one of the issues of biometrics is that they're inherently a fuzzy parameter you can't easily cryptographically prove that your fingerprint is a close match for the one that's stored and you don't want to store the original fingerprint on chain because then anyone can imitate you so the idea might be yes that for example we can have a device okay another cryptographic expert but I guess you know it would be possible that we can have a device that will translate our biometrics not only the fingerprint but we know some neurons or or a heartbeat you know things which are really unique for the people into a kind of key sequence so even solve that I would lose the device the device will we can have an algorithm as which will constantly generate new keys based on on that unique features and use that no set of the keys the sequence of the keys to to provide a fallback or to to identity myself on the blockchain so I mean there must be some key at the end of the day but I think what we are missing here I mean we are talking about the identity and the mass adoption is that this traditional way where we just you know offer one key to the guy that simple doesn't work so what do you think about that where should we go with it because if I will offer it to my mother like oh why what have I how can I deal with it yes within the traditional way yes it's a government whose issuing so if if I my ID card I can always go to government and get a new one and I know make a photo - whatever yeah maybe I can answer on that one so I mean on the end this is exactly for example what the es is 7 to 5 is about is creating a container it's also you pour dat specs it's about creating containers for many keys and for different ways of how you can basically like replace them for examined years 7 to 5 you have different type of keys and the keys you normally use and indirect are the action keys which are not the ones which can overtake your identity or alter your identity and also concerning the biometric data I think probably in the future we will have ways to derive like some uniqueness out of our by my data or you know the simple thing would be I mean you have to do it reliably but I mean the simple thing is you could just put a chip in your hand I mean that's kind of crazy but I mean that's most likely what will happen and I mean the idea here really is to create frameworks to to standardize these kind of like containers of keys and claims and claims can be primary data I mean either it's a co knowledge proof container which then can verify your primary data or it's another provider like the state or whatever you know who saw you once you know and then there is an API call back and you just send the data points there and I say yes or no it's him or not and yeah I mean on the end once we have sanitized the the containers everything else like what gets put into them that's another question so I see it yeah to add what what Fabian says is basically you we're separating the key management with the identity you basically have identity and a way to control your identity if you control with a key with a pair of keys with a multiple multisig account that some keys has are owned by the state or bar whatever you trust or by your biometric derived key somehow will still have control of your identity so all the fallback methods we can build very we have a lot of freedom to build very different kinds of ways to do it as long as the identity remains permanent on and we have a way to control that with different kinds of soft keys and if I can add something to do the question earlier when it's about Candida be leaked right I mean first off today we are in the worst situation like basically you always hand over your data at all times completely no questions asked you know you show the card they can just read everything and or you give credit cards away you know and they make like copies or whatever at the past even and the other part is yeah it's a data can be leaked but in the ideal situation we are not actually giving away the data anymore but we just give me a way the the claim itself like the reference of the claim because in most cases if I'm a bank I just need to know okay he is a human and he has an address in the least cases only in the case when I actually need to access the address I need to know where and probably in the future that your postal service would just send you just send them two IDs and the postal service knows where to route them like I don't need to know exactly where I live in order to sent you a hardware mail or so so I think it's more about like kind of take the claims itself and the under interesting aspect is you could shield claims with many claims so instead of actually having each single property claimed or like different claims for each property you can have like packages right like it's this guy able to access this and then if somebody wishes claim yeah we proved all of these things you never need to know them but if you trust us you know you can give him this or that you know and I think it's kind of like these kind of things will happen to obfuscate that and then the leakage of data is the worst case okay you know he could have acts at that system or not kind of I think it's worth noting too that there's a number of questions you might want to answer with identities and a lot of them fundamentally rely on a central authority like is this person a British citizen is this person over 18 things like that are very difficult to answer authoritative Li unless you point to somebody and say yes they are the authority on who is over 18 or who is a British citizen and biometrics can help protect your authorization your authentication data but I can't prove those things about you one problem I see and correct me if you think it's old is that identity is also consisted of many things that we don't really want to publish so a bank would also want to know if if you have another account with a bank with a depth or something and if you are able to withhold it with like have it another identity account or anything else then how useful is this identity actually but you you can choose what to share with other parties and what to keep secret of course you don't want to share all the identities or the claims that you are holding in your identity but with selective disclosure or some kind of private claims but can be stored of China and you can give the consent to to somebody to see this information but not to everybody so I think the question is also what prevents you from using that selective disclosure to your advantage what if you just declined to disclose all of your bankruptcies for instance maybe I mean somebody wanted here here is I would actually make the difference between identity and reputation right reputation is basically just a summarization of others saying things about me and identity is like what I approve it's kind of a subset of reputation I would say because for example intercepts or an identity space it's about you know there are claims being made and I approved them and I'm kind of conject them also and the reputation system would tell you ok that's that they say you know he's like not valid he like made a bunch of depth or so and then there's the question of exactly how this could be accessible or where this could be stored for some of the ESD 780 could be a reputation system or you could have could it be based on points or I don't know and it's hard to tell I mean right now we have in Germany - ooh file system you know which is just a score you get based on whatever they saw so I think this the two different systems and they have to probably only cross-reference so well it's for example if I'm convicted that's really like hard claim about my identity it's not reputation I'm convicted and probably they are not allowed to give me some services or something I don't know I by law or something so this is part of identity so if if I'm able to actually have this claim attached to my second identity so I'm doing like cybertek or something then in the end some somebody who is serious about it can't really trust anybody any any identity account because maybe I'm actually also convicted but it's just not attached to the one that I provided I would say the difference is here the ones which has a physical person connected with probably one which wants to be at first convicted and second want to be checked against probably not your game identity you know might be a fine person yeah I I think another way to handle this too is to have negative claims so you could be required to release information about the claim by the corrections department that you have no convictions and if you can't present that then that tells you something okay okay I have a question concerning I mean just something we discussed before like separating the access and giving access for with a certain identity you have different providers here and the data itself the data management because the data is still the same so yeah I think it would be interesting to separate both just to you know gather all these different data materialized the the data management in this case gathering all these biometrics or any type of identity personal data did you I just won big on the question that you made before about this leaking of information I think it's very important to distinguish that if you prove me something to me I can publish this proof and this is one thing and the other thing is you publish something to me and I am claiming that you are whatever these two things are different of course I can claim whatever I want but there are systems where you can disclosure some information to me and I cannot reuse this information to the rest of the people because maybe he is giving me zero knowledge proof about that thing I'll or maybe that we know that he knows I just release a random number at that moment so this proof is going to be it can be both things it can be a proof of that that claim exists or that just a random number exists yeah but this proof is gonna be useless because this proof can also means that's just that I knew a random number at the begin at a given none at the given and a given no it's a proof worse that the question is not only about the claim it's about the claim an arena that I know is Pacific not a specific random number so that proof is useless except in the context of where you are but we're proving on to see one that isn't in the small corner of the lake theater have any questions the exact opposite maybe the open question but you mention it in your presentation about phaedo UAF implementation so you mentioned more or less how have a safe store it's about the keys so I wanted to really know because I don't know that the deep inside about Phaedo all of you which solution do you find to really have a safe storage and of the of the keys and what are you willing to achieve in the case of Phaedo instead how our wallet because I didn't get it well so which video you have company are creating federal compliant devices that generate and store private keys private and public keys also and these keys are used to make for the factor authentication okay and which we are what we want what we want to do is to use these devices in the key generated and stored on these devices to sign the transology seven transaction because they are compatible we are seen that by chance the protocol used by fado is compatible by the one used by a firm so we can sign and the thrown transaction with a few doe device or something like that with a few device so that it's a responsibility of the video device to store the key to make them in there to to to leave them in security after if you lose a few devices you lose the keys when you you fall back to the way of recovering key either and well by your implementation or you'll see 725 all the way it's proposed by you port to recover the key when you lose one but the idea here is that the private key itself you never get it it's always managed by the video device and it's a few device that so in the transition it's the same way that it works today with the ledger but it's more standard because using Fido father stands for Feist identity online so it's compatible with fairest on that so it's more interoperable and the question first question a kind of system where you have persistent identifiers pointing to individuals cannot ever be gdpr compliant I mean the fact that I asked this question you can assume what my own answer would be and the second thing that I would like to ask to understand people from in tech that view is that I mean the public block chains aren't they a better fit for actions of legal entities instead of human beings because of gdpr and also the fact that there's often legal there's a mandatory requirement that some actions of legal entities are public and then a public block chain is a good fit well regarding gdpr if so it's true when something is on the public block chain you can't remove it and one of the assessment of gdpr is that somebody can ask to remove his personal information from the provider so problem is a block chain is here there is a thing that has not been entered by GDP is that there is no provider beyond blockchain so if I take my information even if they have been the case I take the information that has been created by your certification Authority I have my information and I take them I put them in the blockchain so I consent to put them on the blockchain here I'm compliant with gdpr because I give my consent to put them on P because I put them on the blockchain myself so I give my concerns and the books and the problem is that they can't be removed so it's it's not handled that whole today by any lawyers we have spoken with lots of lawyers the only thing is that you have a very strong disclaimer that tell the end-user that when this information will be one direction you won't be with a bill at all etc etc but since you have no provider to to go to when we want to remove them it's possible and so it's not completely compliant to gdpr but it's not and compliant at all also because it's just not for and then by the GDP or regulation as it is today so yeah so the question is that as a user maybe I don't want my personal information to be put in the appropriate location and to stay there until the end of time so it's true so today this is the way we propose to do that if we see that nobody wants to put his personal information on the public blockchain even to invest in a CEO thought or something we'll have to find another way maybe we will change to only put your CSN to a certificate number identification number it's more under name that it's linked to the - - it's also a personal information because it's possible - to know the identity from this anonymous number so we have to put something on blog on the blockchain to link the blockchain address with the identity at least there's no but maybe I agree that maybe first-name lastname are are too much for some people yeah sorry last question I apologize if you've already mentioned this but I only caught the last few minutes of the talk prior to this panel but do you have any plans to allow users to monetize their ID for example by selling aspects of the information to market research companies or maybe even governments to fund things like increased participation in voting for example yes since you have control of your data you can exchange your data on something like money some benefits or whatever so basically and you can have us my contract for that as escrow so as long as they put the money on the escrow you exchange your data and you could receive the money back today we has not woken that but definitely yes it's it's it's a possibility and the NDB on it is that as you probably all of you know the famous phrase if if you don't pay you are the product so you it's not true of the plate but your information as a product instead of allowing big companies to make money with this information if your information of values you should be able to get the value directly by yourself and yes ideally the raw business model behind behind this this this is this model of identity yes and I think there is always some but from bringing you to to monetize into to sell your data and I think there is business model for for this and you can also sell anonymous data because some some brand can be interested by consumation data without knowing your name your age your your address so I think that is a use case and it can be very very interesting to have a platform ring you to to sell something that not all that personal information but some kind of consummation that hour why not the good thing with this kind of data which the blockchain technology it's possible to prove that these are real data these are not fake data you are creating just to sell them but they are not related to your real behavior here with we need something something else around the spot consult but it is definitely possible to prove that the data that you sell to the company also right that are you real behavior by you eventually by connecting to a ot2 iot of things you can prove that idea what you have done well that's all we have time for so if you don't give a round of applause to our panelists thank you