Self-Sovereign Identity – What’s Different?
Self-Sovereign Identity: What’s Different
- Wednesday 4F
Convener: Joe Andrieu Notes-taker(s): Dave Sanford
- Tags for the session - technology discussed/ideas considered
Discussion notes, key understandings, outstanding questions, observations, and, if appropriate to this discussion: action items, next steps:
Joe called the session to address how the thinking brought in with distributed ledger technologies has change the aspirational goals (i.e. perception of what is possible) by the IIW community.
Old:
1) Individuals control* how we are correlated across interactions
2) Individuals control* the attributes and claims used to provide services
New?
3) Individuals can selectively assert verifiable and self-asserted claims without dependency upon any central authority
4) That independence is created through free, open standards for cryptographically signed claims (non-repudiability) public ledgers for distributed storage and access control
5) Everybody is a peer, everybody can do all the functions
Much of the below are comments attributed to Joe Andrieu and Christopher Allen, however other unattributed comments are also interspersed:
Chris - If everybody is a peer, everybody is a root (side comment - DNS root, 7 keys which 14 people have, 3 keys needed to make a change)
Joe - What we may have buried is the concept that we don't need a central authority.
Chris - It may be worth a look back at history to understand why and which of Kim Cameron's "Laws of Identity" failed. Need to update the "Self Sovereign Identity Principles"
Joe - Sovereignty is never absolute - it is always a basis for negotiation of boundaries.
6 - Relying party gets to decide which claims to accept
Chris - Administrative IDs will still exist - but there will be alternatives.
May need to go to an issuer to validate a claim or it might be validatable by information (hash, pointer, actual claim detail, ?) on a distributed ledger.
Issues were raised with respect to the right to be forgotten with respect to permanent information on the blockchain. The response was that this is a data minimization problem - with respect to use of the blockchain to be able to verify vs. making information permanently visible. Chris - Up until 50 years ago in West Virginia, two town elders vouching for someone was the basis of claims, this is the historic norm, such 'socially grounded' systems - which we may be returning to in some ways with 'web of trust' type systems.
Chris - Timestamps mean that we can provide continuity over time, which we didn't have as comprehensively in older systems.
