Block 15 IKP Sessions: Proposed Agenda

Working Group(s) IAM, Key Management & Privacy (IKP)
2

IKP Research Pathway: ZKP and Private Biometric Authentication

This post opens an IKP WG research pathway on zero-knowledge proofs applied to biometric authentication, and announces a public working group session to be held between now and September 2026 as the on-ramp to a full session at Block 15.

Why this pathway, why now

Biometric authentication is being deployed at population scale, and the governance question has sharpened: can a system verify that a person is who they claim to be, or that a person is a unique human at all, without that system holding, seeing, or being able to reconstruct the biometric itself?

Zero-knowledge proofs are the most credible answer on the table, and two significant bodies of work have matured in parallel:

  1. World Foundation’s biometric ZKP work, including a demonstrable implementation of ZK verification over iris-derived data, deployed against real uniqueness-at-scale requirements.
  2. The Linux Foundation ZKP working group’s biometric authentication track, developing the open, implementation-neutral treatment of the same problem space.

These two efforts have overlapping problem statements and largely non-overlapping rooms. An IKP WG session putting them side by side is a cheap, high-value convening: BGIN does not need to endorse either approach to be the venue where they are examined together.

The public session (before September)

Format: open IKP working group call, 90 minutes.

Proposed agenda:

  1. World Foundation: biometric ZKP demo and architecture walkthrough (25 min)
  2. Linux Foundation ZKP WG: biometric work, scope and current drafts (25 min)
  3. Moderated comparison: threat models, trust assumptions, revocation, template protection, what each proof actually proves (25 min)
  4. Open discussion and Block 15 scoping (15 min)

Timing: targeting late July or August, ahead of GDC (1 to 3 September, Geneva) so that findings can inform any relevant Geneva conversations, and well ahead of Block 15 (15 to 16 October, Washington, DC), where this pathway gets a dedicated session.

Research questions for the pathway

These are the questions I propose the pathway carry through the public session, Block 15, and any resulting deliverable:

  1. What is being proven? In each system, what statement does the ZKP actually attest: uniqueness, liveness, template match, enrolment integrity, or some composition? Where does the trust boundary sit relative to the proof?
  2. Where does the biometric live? Device, custodial enclave, distributed storage, or nowhere (derived and discarded)? What are the reconstruction risks in each posture?
  3. Revocation and compromise. Biometrics cannot be rotated. What does revocation mean for a biometric-bound credential, and how do the two approaches handle template or key compromise?
  4. Proof of personhood adjacency. How do these systems relate to personhood credentials in agentic AI contexts (see the parallel IKP proof of personhood track)? Is biometric ZKP a personhood primitive, an authentication primitive, or both, and does conflating the two create governance risk?
  5. Standards surface. Which elements are ready for standardisation, and in which venue: ISO, W3C, IETF, the LF ZKP WG itself?
  6. Regulatory posture. How do these architectures map against biometric data protection regimes (GDPR special category data, state biometric statutes, emerging digital identity frameworks)?

Relationship to other IKP work

This pathway sits alongside, and deliberately does not merge with, the proof of personhood paper track. Biometric ZKP is one candidate mechanism for personhood credentials, but the pathway examines it first as an authentication and key management question, which is the IKP remit. Where the two tracks converge, they will do so explicitly at Block 15.

Asks

  1. Speakers: confirmed contacts at World Foundation and the LF ZKP WG willing to present at the public session; if you hold those relationships, reply here or reach out directly.
  2. A co-moderator for the comparison segment, ideally someone with applied biometrics or template-protection background.
  3. Date preferences: indicate late July vs. August availability in this thread.
  4. Additional research questions to add to the list above before it hardens into the session brief.