‹ BackHN Continuity

Thread

Identity Management for Agentic AI [pdf] (2025)

82 points · 28 comments · cgeier

  1. juanre · · focus · HN ↗
    I wrote a solution for this, and it has been working across several organizations for some months now. I run it as a public service at <a href="https:&#x2F;&#x2F;awid.ai" rel="nofollow">https:&#x2F;&#x2F;awid.ai but I would love to offload it to a foundation.

    - OSS (<a href="https:&#x2F;&#x2F;github.com&#x2F;awebai&#x2F;aweb&#x2F;tree&#x2F;main&#x2F;awid" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;awebai&#x2F;aweb&#x2F;tree&#x2F;main&#x2F;awid)

    - Trust rooted in the DNS.

    - Multiple registries supported.

    - did, verifiable stable identities.

    - Certificate-based teams.

    It is being a very interesting project so far. It&#x27;s at the core of my <a href="https:&#x2F;&#x2F;aweb.ai" rel="nofollow">https:&#x2F;&#x2F;aweb.ai which enables agents to communicate globally, and it makes it possible to build really simple agent-first tools where auth is just a matter of validating a certificate.

    1. brw · · focus · HN ↗
      what does this offer over did:web[0]? or if a central registry + op log is required something like did:plc[1] which came from atproto&#x2F;Bluesky[2]?

      [0] <a href="https:&#x2F;&#x2F;w3c-ccg.github.io&#x2F;did-method-web&#x2F;" rel="nofollow">https:&#x2F;&#x2F;w3c-ccg.github.io&#x2F;did-method-web&#x2F;

      [1] <a href="https:&#x2F;&#x2F;plc.directory" rel="nofollow">https:&#x2F;&#x2F;plc.directory

      [2] <a href="https:&#x2F;&#x2F;atproto.com&#x2F;specs&#x2F;did" rel="nofollow">https:&#x2F;&#x2F;atproto.com&#x2F;specs&#x2F;did

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.