Coordinate with PING on testable pairwise identifier derivation primitives

[amir-hameed-mir]

Issue Description

We have developed comprehensive test vectors demonstrating testable pairwise identifier derivation with formal privacy properties. This addresses a gap in W3C's horizontal privacy guidance where specifications recommend "pairwise identifiers" but lack testable definitions for derivation, unlinkability, and recovery.

Our Work:

• Created 15 privacy test vectors showing deterministic pairwise DID derivation
• Formal adversary models with quantified security bounds (≤2⁻¹²⁸ advantage)
• Cryptographic properties: unlinkability, compartmentalization, recovery
• 230+ total test vectors across 9 protocol categories

The Gap:
W3C specs (DID Core, VC, etc.) mention "pairwise identifiers" but don't define:

1. How to derive them cryptographically
2. What unlinkability properties to guarantee
3. How to recover without reintroducing correlation
4. What adversary models apply

Next Steps for UDNA Community

1. Review our privacy test vectors: /udna-test-vectors
2. Discuss coordination approach with W3C Privacy Interest Group (PING)
3. Plan presentation to PING showing our reference implementation
4. Define path to incorporate testable privacy primitives into UDNA specification

Questions for Discussion

1. How should we engage with PING? (Email, meeting presentation, joint document?)
2. What's the best way to position our work as helpful to W3C privacy guidance?
3. Should we propose a Horizontal Note or contribute to existing PING work?
4. How do we ensure UDNA's privacy approach aligns with W3C-wide guidance?

Action Items

• Review privacy test vectors
• Draft email to PING chairs
• Prepare 10-minute presentation for PING meeting
• Identify affected W3C specifications
• Plan specification updates based on PING feedback

Reference: Full test vectors at /udna-test-vectors/