Day 2: Jason Law (Evernym)

Transcript

[Music] next up is our final speaker on the identity management block it is Jason law co-founder CEO and chairman of blockchain identity solution provider Evernham Jason okay okay is that what I look like from behind thanks I know it's X after lunch and we're maybe a little groggy but I want to share a few things I want to talk about a model for self sovereign identity now we've heard a bit

about zero knowledge proof and about blockchain about tamper evidence about consensus there's a number of elements that go into any kind of a solution and I'm kind of lucky because a lot of people have talked about things that I normally would talk about zero knowledge proof for example and and dr. Buchanan and facili Suvorov have mentioned those things but you see I have some colors up here I'm carrying a couple of roles here I I started ever I was a co-founder of Evernham but then sovereign foundation is something that kind of grew out of what we were doing and I want to talk a little bit about that there's some

confusion sometimes about what is sovereign and what is Indy Indy is the code base that sovereign contributed to the hyper ledger project under Linux Linux Foundation and Evernham has been helping build that since 2015 so just want to make sure the separation was understood and I guess I can look here I'm gonna skip through a couple of slides here we get a number of different words thrown around when we talk about identity and blockchain and they're mostly the same the standard I think that we're trying to achieve is something that probably follows this definition a little more closely and

that is that it's it's a lifetime portable identity that is fully controlled by you and it cannot be taken away simply anchoring credentials to a ledger may not be enough because if you have to go through a system that can can stop you from using it then then you it's not really self sovereign what's exciting is a few years ago when we started talking about self sovereign identity and there were others talking about it there were not very many of us and we felt like weirdos we we would we bring things up and and we'd have people looking at us a little they squint at us and blink at us a

couple of times but it's it's exciting to see that it's kind of taking off and to hear the the projections in the space are it's just it's fortuitous so I want to talk about a model and there's three different dimensions to this model and I have pictures so it'll be a little more clear when I show them there's three dimensions and the first is relationships and let's say that I've met someone here and I and I've talked to them about maybe where I'm from and we've exchanged some stories they know more about me but I met someone over here and maybe someone over here doesn't know the same stories in real life we don't have one identity internally we might think of

ourselves as having one identity but to others we have multiple facets and so it's important to think when we build systems that the system should model what happens in real life and so one of the first things we kind of realized was that having one identifier was actually very harmful and you'll kind of see how we can we can have a strong identity without having a correlated will identifier and correlation handles in general attributes we all have attributes we all have things that we know about ourselves that we have other people say about us we want to be able

to share those the way we want to and agents I think facility mentioned agents um this has become a critical part of infrastructure and we don't realize it but we all use agents all the time but we we don't think about it and so we're going to bring that up to the surface here the very first one we're going to talk about is relationships I talked about identity with the capital I versus identity with a little eye and identity what the little eye is really contextual Who I am to my children in fact Who I am to each of my children is very different even though they may be in the same room with me for a lot of

the time that I'm with them who I am to to my my family to my bank to my employer are all very different and we need to treat those as different identities the second dimension is attributes this is information about me that I hold that I can share the things that people know about me and the last one is agents and you see the direction of these you can kind of imagine if these are mutually orthogonal dimensions then then they can create you can imagine them in a three dimensional space which is going to where my head goes but before I show you that I want to show that I can't digitally sign

anything you give me a piece of paper and a pen and I'll sign something no problem but I can't I can't compute a signing algorithm I can't store a private key in my head I couldn't verify one and so we all talk about Alice and Bob having keys but they don't have any keys what Alice and Bob have are something like this that holds a key that computes the computes a signature or verify the signature and so making these first-class citizens in an identity model was very important and then the last one is a cloud agent which

is an interesting concept right now when we use cloud services we're using somebody else's services they put an interface on it for us when I when I send a message usually I send a message that goes through a server that someone else owns they promise not to look or they promise to secure it yeah there can be some mechanisms for for me ensuring that that is is secure going through their servers but I'm still going through their servers and having a a way that I have my own ability to control what happens in the cloud is is quite intriguing and there's some neat use cases for it so here's the the XY and z space the

three-dimensional space and we're going to talk a little bit about the intersection of the planes between these and this is where it gets kind of interesting so Vasily talked about this a little bit the concept of me being able to well let me talk about just a general high-level and then we'll go into the details this plane is who knows what about me the next plane is which proxy or which agent represents me and in which context and how and lastly which proxy can share what information about me so you can see

each of the planes kind of pull out the in this case relationships is is pulled out but the combination of the three create a fairly rich landscape for control for an individual when I say individual I mean individuals or entities companies whoever needs to maintain a self sovereign identity so I want to dig in to this one a little bit and I think this will make it more clear these lines up up above represent a different relationship that I have and I'm employed by acne and acne knows some things about me it knows my name it knows a lot of things about me but I'm

simplifying it knows my name my role in the company with salary my start date and this example is I want to go get a loan so my employer can give me a credential and a credential is a collection of attributes that are signed by that by that employer and but they're signed in a way that I can prove that I have a signature and prove the contents of that credential without actually revealing the signature because a signature like a hash is universally unique and is a correlation handle so I want to get a loan with UVs and I get to using zero

knowledge proof prove my name and my salary without actually needing to reveal the credential itself and so that's why I've painted these with the with a different symbol so let's talk about another another plane here here I've got UBS Bank I want to I'm actually maybe a customer of UBS now and that line represents the intersection of all of my agents with UBS now not all of my agents need to be part of the relationship with UBS but let's say for example my phone is so my phone has a key and this is how you BS knows it's me because I can sign

my communications with little a the private key and UBS knows that it's me and let's say I also want to use my notebook we don't share private keys across devices so I have a separate key for that one and I have a separate key for perhaps cloudy agent because maybe when I have a I have a mobile phone in a notebook because I have in sovereign you have a pairwise relationship you have an asynchronous secure communication channel between the two parties which means that I can get messages from UBS Bank on the same Channel it's a bi-directional channel but if I'm using my mobile phone in my notebook and my

mobile phone is maybe out of battery in my notebook is switched off then I'm I'm out of luck and so maybe I have a cloud agent that represents me with the bank for certain things maybe it's just to allow me to get messages when I'm offline I think it can be much more than that but this is a simple example so I think Vassili also mentioned micro Ledger's micro Ledger's are interesting because they're very small tamper-evident Ledger's with with a consensus of one if you will only one author on this ledger which is me and I have one ledger micro ledger per relationship and here you can

see that I am putting these keys a B and C are on my micro ledger and I share this with you B S UVs gets a copy of it and I have a way to communicate that micro ledger tbs so UVs knows if I change my key and then I added key C and if you can see key C actually has a message procs it's actually I'm telling you bs don't trust this agent or this key for anything but message proxy so send that send a message to that but don't don't actually trust it for anything else and with my banking app I may have a higher level protocol that allows me to

authorize payments and so maybe I say that B because it's my notebook in my notebook and leave it in my office all the time or my kids use it then maybe I don't trust it as much as I would maybe my mobile phone and so I can have an ability to say hey UBS beacon authorize up to $100 and hey you can anything over $100 in fact yeah anything over anything period I'm giving kind of a little bit of a complex example to show that you can do fine-grained authorization of agents for a relationship and you can

imagine any kind of protocol could run on top of this and could could leverage this as kind of a dynamic self sovereign access control list for relationship okay this one's better so let's talk about the last one I'll go quickly through this and I think we're doing good on time so I've got my four and four agents here a mobile phone a notebook and iPad and a cloud agent I labeled these W X Y Z and if you remember agents and attributes plane is all about which agents are allowed to share what data

about me and there's two layers to this first layer is if I have a phone that doesn't that has my credentials on it and I don't want another device to be able to prove something from that credential then I just I don't share it with that other device that's one layer and there's there's a simple way to do that but the second layer which I think is a little more interesting is about recovery when I go self sovereign I put I put keys on my device that that allow me to prove things and I no longer have some of the protections and so I have to have a really good decentralized key

management system I've got to have a way to be able to let someone else trust that it's really me and not someone who's taking my phone so having the ability to have some sophisticated policy around you know recovery and key rotation and things like that is important now of course we don't want to put we don't want to burden a user with all of that all of that information and make them have to make those hard decisions so it's up to the apps and this is going to be something that's really important over the next few years is the apps need to help the user be smart about security and about recovery and and everything related to that so here with agents and attributes you have

a policy and in soverign you can do this in several different ways and you could do this in aetherium too but in soverign we have what's equivalent to a smart contract that allows you to build a policy or your agents and you have one per identity with a capital I so one entity or one person would have one of these and you can see that this is authorizing w x and y which are my edge agents to prove something and so now when I when I issue a zero knowledge proof of my credentials I'm actually showing I'm actually proving that I'm doing it from

one of these devices and I'm doing that in zero knowledge so that it can be correlated to a particular identity on a ledger number two I can set a policy something more sophisticated like if I'm going to grant provisioning of new agents I'm going to need two of those three devices to come together and do some sort of a multi-sig and this is just an example just to kind of illustrate what's possible with this model I add a third one Z which is a cloud agent remember cloud agents are not our agents that are not in your control they still have keys they're still authorized by you but but we might trust them less so we call them maybe semi

trusted and so but I would trust Z to participate in a revocation so I lost my phone a few months ago I was in Corona ghen Netherlands and I flew to South End London and I left it in the back of my of my seat back and I spent I don't know two three hours cumulatively calling the airports calling the airline I never recovered it and of course it's unlocked because I'm here and I'm using SIM cards in Europe and I was I was very frustrated but but if if I'm using that for my identity then I want the ability

to use my other devices or even other people to help me revoke that so that someone can't impersonate me with that device and I have to do that without being able to reach the device because a smart person is going to to turn off the network they're going to take their time they're gonna hack it and then they're gonna be very selective about how they use it so this kind of a policy allows you to have comfort in that scenario and just one note about defuse trust what makes blockchain powerful is the notion of defuse trust that it takes a number of people to cooperate to collude in fact too many people to collude to create a

problem and so pushing defuse trust out to the edge defuse trust can live everywhere and pushing it to the edge is a very powerful way to do it and so in an in an EM event type scenario it's it's helpful there and then of course I lost my phone so I have a new transaction here to revoke W so W is no longer able to prove on my behalf so I probably went a little deep on this but I just wanted to illustrate that the model needs to support all three dimensions and all three planes and this is a work in process this feels good this has been a model we've been working against for a long time but if this

sounds like like it's wrong or there's something missing I'd like to hear about it and I have my contact information at the bottom at the last slide I wanted I want to talk more about it because i i i i i want to challenge what we've done and i'm happy with this but i want to challenge it just one more note on agents it's we've talked a lot about at this conference about IOT and things like self-driving cars agents can be anything that we want you can think of agents as being anything my home security system is an

agent my doorbell can be an agent my car my wife can be an agent or at least one of her keys can be an agent for authorizing high-risk transactions she says it's for my protection but i really know she's keeping an eye on my my financial transactions my friend for key recovery my friend knows me and I can call him and I can talk to him and he can he can judge whether or not it's really me much better than maybe a simple password verification my attorney I could authorize my attorney to do certain

things or my realtor and so you can imagine this the use cases that can sit on top of a model like this I my my partner's video is common he he would say things like your agent is like your Jarvis and you know I love this concept I love the idea of of having something that's kind of watching out for me that doesn't work for someone else that works for me and can help me participate in collective intelligence in a commercial way without exposing Who I am and all of my details I love that idea the project's been going and I skipped

over some slides up at the top it's kind of it's taken off the last 18 months have been crazy I want to thank a few people that have kind of contributed to this and then tell you how you could participate Cooper Valley Association of course hype religion Linux Foundation I love what these guys did they they pushed us because we are taking a strong stance on privacy they really wanted to get our project in and I'm really glad we did because they've offered a lot of support I want a shout out to Yan camenisch he's here in Zurich and his team at IBM IBM is as a founding steward of sovereign and and Yuans work has been instrumental

absolutely instrumental because it's paved the way for the privacy that we needed to have in order to do something like this if we didn't have that work we wouldn't have been able to do this so I want to thank him and his team thank Vasily Suvarov for for asking the really hard questions SiC but he's here Iran they're a steward in there doing great work and grateful that Cisco was able to announce yesterday that there are steward and also just uh we're doing a few things Alex Brooke shot with us as I meet up we're I want to talk a little bit about kind of a broader initiative around SSI

I'll do that in just a second if you're interested go to hyper ledger in D also sovereign you can participate in a working group there's technical governance board than I chair that the you know if you if you'd like to participate let me know w3c verifiable claims working group is in fact let me just talk about that we have a pull request right now that's going into this back it's not a spec yet it's not finished yet but we think this year we'll have it but the pull request is turning verifiable credentials into zero

knowledge proof friendly verifiable credentials it's exciting to see that happen because that's been it's been a lot of work to kind of create awareness and get people really thinking about why zero knowledge proof SAR so important in blockchain applications decentralize it any foundation participate on that and that's a good place you ports a member and and Microsoft and some other big ones to remember that SSI meetup is brand new we're trying to I'm helping try to get this started a bit but it's really kind of ledger agnostic let's find people that are doing stuff in the space and let's learn from each

other there's so much tremendous opportunity in the space that it's it's you got to have an abundance mentality that there's so much to to to do and and there's so much work to do and they're starting to be convergence on the right kinds of ideas the good ideas and I think we all benefit when we come together and we learn from each other so I'm excited that this is happening and then this is it's a little bit early but we're starting to talk about a common protocol something like SMTP for self sovereign identity I think this is a pretty positive notion that a number of

the players including you port and Varys won and the DI F are people that the di F are all starting to talk about a common protocol so this really helps us all and and it's exciting to see that happening so I think I'm a little bit early and I was kind of running fast but here's my info you can send you can reach out to me on LinkedIn or my email address and then I won't respond and then two weeks later I'll finally respond maybe I'll give you the ever name calm and their sovereign org addresses but would love to have your participation Thanks

[Applause]