Identity Hub: the next version of the People Chain for the benefit of the whole Polkadot ecosystem

[ntn_x2]

Hello people! Antonio from KILT here.

Over the years, we at KILT have noticed that no identity solution fits all use cases. We found out that some people like to be publicly recognised for their achievements and build online trust and reputation. Some people prefer to stay as private as possible, but still rely on technology and digital solutions to carry out their everyday tasks. Some use cases only require people to show proof of age and nothing more. Some other cases require them to prove to be humans. Some use cases do not involve people but objects, digital systems, and nowadays, a secure identity solution must also be available to them as well. Some other entities may not even be able to use cryptography, yet they must be certified or attested.

That’s why we have spent the last few months thinking: how can we make it easier for Polkadot as a whole, as an ecosystem, to be the de-facto solution when it comes to deploying identity solutions? The heterogeneous nature of Polkadot makes it the perfect candidate for different projects to tailor different use cases, both inside but especially outside the ecosystem.

One one hand, such heterogeneity is great as it fosters bottom-up innovation and allows different projects to focus on problems they deem important to solve. On the other hand, what usually comes out of it is pure chaos™. That is what we have noticed over the years. Each project tries to establish itself as the one-fits-all identity solution, making it extremely hard to foster collaboration between all these projects such that the total value the Polkadot ecosystem created is the sum of the value that each project creates.

Furthermore, we think that moving away from a crypto-first and blockchain-first mindset, especially in the identity space, would allow more industry players to explore the possibilities that Polkadot, as an ecosystem, unlocks. That is why we started looking at identity technologies that have enabled giants such as Google and Facebook to become who they are and try to bring them inside Polkadot to see if we can build on that and improve the status quo for the ultimate benefit of the larger community.

What we found out is that the most value is created by following two simple principles: 1. delegation of responsibilities, such that each project focuses on what it can do best, and 2. linking and reusing data, such that the value a project or player creates can be consumed and at the same time enriched by other players.

Probably the best example of the first point is the OpenID decentralised authentication protocol, which spares websites and applications from the need to roll out their own authentication system and rely on partners that provide identity management solutions as their main service. At KILT, we have tried to bring that inside the Polkadot space with the Decentralised Identity Provider (DIP), which mirrors OpenID-like functionalities for cross-chain identity sharing. We gave a talk at the Sub0 in Bangkok. Feel free to check it out.

The second point, linking and reusing data, is what we are proposing today. Because we acknowledge there is no single identity solution that can solve all the problems, we are suggesting a framework for identities to be standardised in a way that each project can at the same time, solve the problems it can solve best and contribute to the overall value Polkadot provides. We propose to model identities after RDF triples, which is the most common data structure used for linked data. Our proposed solution is flexible and generic since it does not favour any specific solution. Instead, it invites all identity solutions to contribute their value creation to the Polkadot ecosystem without sacrificing innovation speed.

We at KILT are extremely thrilled to present this proposal, and we hope and think the Polkadot community will be as well. Here is our high-level proposal and vision for Identity Hub, which we see as a natural evolution of the People system parachain: https://docs.google.com/document/d/1F22XI2naMgbQC8AuwI-ni-yp6IMEvlqc6NnE4eMVphI. We are hoping for an extensive community engagement and are ready to answer any technical and non-technical questions at any time.

Cheers,
Antonio and the KILT team

[nino_apillon]

Is this new approach systematically above DID W3C standard, so it is like a solution that encapsulates an option to use DIDs as well as other “identity” models?

[alice_und_bob]

Using RDF as described in the proposal is an excellent idea.

Do you think that the age of a claim plays a role and should somehow be represented?

[ntn_x2]

This is not a specific solution, but rather a framework that is generic enough to encompass different solutions. DIDs (and VCs, for what matters) are nothing more than a standardised container format for linked data (serialised as JSON-LD). So they would be 100% compatible with RDF representation. All the RDF requires for subjects is to be URIs. While DIDs are URIs by design, other ways of identifying subjects (i.e., accounts, other identity claims, parachains) might require some “specification” as to how they could be represented as URI strings.

TL;DR: This framework sits above any specific identity system, and would allow any identity primitive to be used, yes.

[ntn_x2]

I think the less we deviate from the RDF standard, the more we can rely on existing technologies, such as SPARQL as a querying language, as mentioned in the document. The age of a claim itself could be on its own a claim, so I don’t see a need to add it as a special field to a claim. The only special field I would envision, given the nature of Polkadot, is the source (meaning the parachain) of a claim, which is an implementation detail. Everything else should fit within the RDF framework, which is very generic and allows for all these properties to be expressed in a nice and concise way.