Skip to main content
Open questionOpen question

Does any published standard require an agent to hold an identity distinct from the person it acts for?

No. The Model Context Protocol authorization specification requires the strongest scoping control available, audience bound tokens under the OAuth resource indicators standard, yet still defines the client as acting on behalf of a resource owner. The standards body work on agent identity is an initial public draft concept paper, and the internet drafts competing to define it have produced no published standard.

The verdict

Open question

No settled answer exists. The ledger poses the question, links the live debate, and does not answer it.

Key facts

What the sources say

Record ID
AIL-2026-0036
Kind
Open question
Jurisdiction
Global
Last verified
Added
  • The Model Context Protocol authorization specification states that clients must implement resource indicators for OAuth so that a token names the resource it is for.
  • The same specification defines the client as acting on behalf of a resource owner and does not require the agent to be a principal distinct from the human.
  • Authorization is optional for implementations of that protocol overall, so a deployment can hold none of it.
  • The United States standards body's contribution on software and AI agent identity and authorization is an initial public draft concept paper seeking comment.
  • Several competing individual internet drafts propose agent authentication and authorization schemes, and none has become a published standard.

Dimension by dimension

3 dimensions, each one stated, silent or open

Identity, Authorization, Delegation. Stated means the document you can open below says it; silent means the ledger read the document and it does not.

IdentitySilent
No published standard requires an agent to be a principal separate from the human it acts for, which is why every incident in this ledger reads in logs as the human.Model Context Protocol specification, authorization, primary source, 18 June 2025.
AuthorizationStated
Audience bound tokens are a real and mandatory control in one widely deployed protocol, and they bound the token rather than the actor.Model Context Protocol specification, authorization, primary source, 18 June 2025.
DelegationOpen
The work that would define an attestable chain from a human to an agent is at draft stage in more than one body.IETF datatracker, draft on authentication and authorization of AI agent interactions, primary source, 6 July 2026.

What it changes

For a team deploying an agent

You cannot wait for the standard, and you do not need to. Give each agent its own service account or connected application today, so its actions are separable in your logs, its permissions can be narrowed without touching a person's, and it can be revoked without locking out an employee. That is the single change that makes every other control on this list enforceable.

Sources

What this record was verified against

  1. Model Context Protocol specification, authorizationPrimary · 18 June 2025
  2. NIST NCCoE, accelerating the adoption of software and AI agent identity, initial public draftPrimary · 5 February 2026
  3. IETF datatracker, draft on authentication and authorization of AI agent interactionsPrimary · 6 July 2026

Related

Records that sit beside this one

Cite this record

Free to reuse under CC BY 4.0, with attribution. The record ID AIL-2026-0036 is permanent and is never reused.

In a sentence
According to the GAGE Agent Incident Ledger (as of 15 September 2026), does any published standard require an agent to hold an identity distinct from the person it acts for?.
APA
GAGE (Global Academy of Generative-AI Education). (2026). Does any published standard require an agent to hold an identity distinct from the person it acts for?. Agent Incident Ledger. Retrieved 15 September 2026, from https://www.gage.academy/tools/agent-incident-ledger/records/AIL-2026-0036-no-standard-requires-an-agent-to-hold-its-own-identity
MLA
"Does any published standard require an agent to hold an identity distinct from the person it acts for?." Agent Incident Ledger, GAGE (Global Academy of Generative-AI Education), 15 September 2026, https://www.gage.academy/tools/agent-incident-ledger/records/AIL-2026-0036-no-standard-requires-an-agent-to-hold-its-own-identity.
Chicago
GAGE (Global Academy of Generative-AI Education). "Does any published standard require an agent to hold an identity distinct from the person it acts for?." Agent Incident Ledger. Last modified 15 September 2026. https://www.gage.academy/tools/agent-incident-ledger/records/AIL-2026-0036-no-standard-requires-an-agent-to-hold-its-own-identity.
Permalink
https://www.gage.academy/tools/agent-incident-ledger/records/AIL-2026-0036-no-standard-requires-an-agent-to-hold-its-own-identity

Last updated . Every record re verified . The ledger is checked weekly, every Monday, and the same day for any vendor disclosure.

Back to the full ledger, or every record for Global and every open question record.

GAGE briefings tell you which AI regulation deadlines are coming, what they actually require of you, and when a program opens.

Powered by GAGE, free EU AI Act tools from the Global Academy of Generative-AI Education
CallEmailHelp