Skip to content
IAM Insights

Explainers · Glossary

WIMSE

Workload Identity in Multi System Environments

By Boston Identity · Explained · 7 min read

The short answer

WIMSE is an IETF working group, and its architecture draft, for how software services prove who they are to each other with key-bound credentials and carry each request's security context across gateways and trust domains.

What it is
An architecture
Defined by
IETF WIMSE working group, architecture draft -08 (Informational)
Who it's for
Teams securing service-to-service and agent traffic
In our briefings
82By month of the source: 2025-12: 0, 2026-01: 0, 2026-02: 4, 2026-03: 3, 2026-04: 3, 2026-05: 3, 2026-06: 3, 2026-07: 13, 2026-08: 4, 2026-09: 15, 2026-10: 25, 2026-11: 22by month, last 12
01

What is WIMSE?

WIMSE stands for Workload Identity in Multi System Environments. It is an Internet Engineering Task Force (IETF) working group, and its architecture document is a draft, not yet a standard: revision -08, published 6 July 2026, with an intended status of Informational.S1 A workload, in the draft's terms, is software that runs for a specific purpose, such as a microservice, a container, a virtual machine or a serverless function.§2

When one workload calls another, the receiver needs to know which workload is calling. Often it also needs the security context of the request: who the user is, or what processing has already happened.§2, §3.4.6 The draft builds its answer from three parts: a trust domain, a workload identifier and identity credentials.§3.1

A trust domain is a group of systems that share one set of security controls and policies, and it should be named by a domain name the organization owns.§3.1.1 The credential is either an X.509 certificate or a JSON Web Token (JWT) called a Workload Identity Token (WIT). Each comes paired with a private key that is kept secret and used only to prove the presenter holds the credential.§3.1.3 Deployments may also use plain bearer tokens, which work for whoever holds them, to interoperate with older systems that cannot handle key-bound credentials.§2

Not the same as SPIFFE, whose deployments have issued workload certificates for years.1 The WIMSE architecture names the SPIFFE identifier only as one example of a format that fits its rules.§3.1.2

02

How does WIMSE work?

The draft explains itself through example deployments. We follow one request through them: an outside client, a gateway at the edge of the trust domain, a credential and token service, two internal workloads, and a service in another trust domain.§3.2

Figure steps
  1. The credential service equips each workload
  2. The client sends its request to the gateway
  3. The gateway gets a context token
  4. The gateway forwards context, not the access token
  5. Workload 1 calls Workload 2 and proves who it is
  6. Workload 2 trades its credential to leave the trust domain
Fig. 1

Each workload gets a key-bound credential first; the gateway then trades the client's access token for a context token that travels with the request, and Workload 2 swaps its credential for an access token before calling another trust domain.

Drawn from the WIMSE architecture draft -08, sections 3.2 and 3.3

Swipe for the full figure →

  1. The credential service equips each workload

    Early in its life, each workload typically gets its identity credential from a credential service tied to its trust domain.§3.2.1 The credential carries a workload identifier, a name that is unique inside that trust domain.§3.1.2 The draft leaves the delivery method open: direct provisioning, a bootstrap credential, attestation (evidence about the platform the workload runs on), or a helper agent on the same host.§3.4.1

  2. The client sends its request to the gateway

    An outside client sends a request to the gateway over HTTPS.§3.2.2 The client may prove who it is with mutual TLS, where both sides present certificates, or with a credential such as an OAuth access token obtained earlier.§3.2.2

  3. The gateway gets a context token

    The gateway asks a token service for a security context token, which holds information about whoever is accessing the system.§3.2.2 One way to do this is to send the client's access token to a token exchange endpoint. The context usually matters only inside the system, so it is not returned to the client.§3.2.2

  4. The gateway forwards context, not the access token

    The gateway passes the request and the context token to the right workload.§3.2.2 The client's access token is not usually forwarded, because it is meant only for external access. The workload uses the context in its authorization policy, and it denies requests that rely on the context if no context token arrives.§3.2.2

  5. Workload 1 calls Workload 2 and proves who it is

    When Workload 1 needs Workload 2, it sends the context along.§3.2.2 Workload 2 may have a policy requiring Workload 1 to authenticate.§3.2.2 Workload 1 can do so with its credential, for example through mutual TLS or with a Workload Identity Token plus proof that it holds the private key.§3.3 Workload 2 then makes a separate authorization decision, and the component enforcing it may consult a policy decision point.§3.3

  6. Workload 2 trades its credential to leave the trust domain

    If Workload 2 needs a service in another trust domain, it first authenticates to the token service with its workload credential and gets a token it can use outside.§3.2.3 It uses that token to obtain an access token for the external service, then presents the access token to that service.§3.2.3

03

Why does WIMSE matter now?

A new framework for agent identity puts WIMSE at its base. Engineers from six companies, among them AWS, OpenAI, Okta and Ping Identity, wrote a framework in which an agent is a workload with a WIMSE identifier that gets access through OAuth 2.0.2 The WIMSE working group adopted that draft on 9 September 2026, and it continues as draft-ietf-wimse-aims-00.2 A survey of ten standards documents for agent identity graded only WIMSE with SPIFFE as available for production, though the grades are the authors' own judgement.3

Readiness is uneven. By the implementers' own reports, the certificate form of the WIMSE credential is in production at SPIFFE, Defakto Security and Teleport, while the token form is at beta, alpha and research status at those same three.1 Uber, which built its own delegation chain, tracks the working group for alignment rather than running its drafts.4 Across vendors, agent identifiers remain split between JWT-based schemes, workload-identity standards and proprietary formats.5

The architecture treats an AI agent as a delegated workload, not a new kind of identity.6 For agents calling agents it says each hop must scope and re-bind the security context, a rule stronger than OAuth token exchange, which leaves narrowing to the server.7

04

Where does it fit?

Compared withHow WIMSE differs
SPIFFESPIFFE deployments already issue workload certificates. In WIMSE, a SPIFFE identifier is one example of a format that fits the identifier rules,§3.1.2 and the WIMSE certificate is compatible with SPIFFE's X.509 identity document.1
Transaction TokensA transaction token carries one request's context between internal services. WIMSE allows context to travel as a transaction token, one option among several,§2, §3.2.2 and the transaction-token draft leaves the workload's own authentication to WIMSE mechanisms.8
OAuth Token ExchangeOAuth Token Exchange (RFC 8693) trades one token for another. The WIMSE architecture points to it when a workload must act for someone else: the workload authenticates to a token service and receives a token for delegation or impersonation.§3.4.7
WIMSE Workload Proof TokenThe Workload Proof Token is one WIMSE mechanism, not the architecture. It is a signed JWT that proves the caller holds the key in its credential.9
05

What goes wrong?

  • Authentication is not authorization. A valid credential shows the workload controls the key and was issued under a trusted authority; it does not mean the workload may reach every resource in its trust domain.§4.5
  • Gateways blur who is calling. TLS protects traffic only between each adjacent pair of endpoints, so a destination must not assume that an identity an intermediary authenticated is the caller's, unless protocol or policy establishes that link.S1
  • Bearer tokens leak. A bearer credential exposed in traffic or logs can let an attacker impersonate a workload or move laterally; key-bound credentials are typically less exposed because the key is not disclosed.§4.4
  • Logs can leak secrets. The draft says each authenticated request must leave a verifiable trace, and audit logs must not contain secrets such as bearer tokens or private keys.§3.4.5
  • Revocation is weak. The WIMSE credentials draft says JWT credentials are not generally revocable on demand, so short expiry is the main defence for a stolen token and key.10
  • Some parts are unfinished. The cross-boundary section still carries a to-do note about aligning with OAuth's cross-domain work.§3.4.10 The two protections the draft names for agent chains, cryptographic binding of delegation tokens and attestation of what an intermediary did, are mentioned but not specified.7
06

What should a security leader ask?

  1. 01

    Which of our services still authenticate to each other with bearer tokens or static API keys, and what would it take to bind them to keys?

  2. 02

    Where do our gateways, load balancers or service meshes end TLS, and how does the receiving service learn who the original caller was?

  3. 03

    When one of our agents calls another agent or service, which component narrows its authority at that hop, and where is that recorded?

  4. 04

    How fast can we stop accepting a stolen workload token, given that JWT credentials are not generally revocable on demand?

  5. 05

    Does our vendor support the WIMSE token form in production, or only the certificate form?

07

Sources

›Evidence: 10 claims from 8 briefings
  1. 1BI-FIND-2026-0375 · The IETF's workload credential binds one key to one name and says nothing else about the software holding itHalf of this draft is already running and the other half is not, and the draft says so. The Workload Identity Certificate is an ordinary X.509 certificate with the workload identifier in exactly one URI SubjectAltName, which the implementation section calls 'fully compatible with the X509-SVID' that SPIFFE deployments have issued for years; SPIFFE deployments simply use their spiffe URIs as the subject. The new piece is the JWT form, and the self-reported status reads beta at SPIFFE, alpha at Defakto Security and research at Teleport, against production for the certificate at all three. The IETF is standardizing the certificate the industry already has and attaching a token nobody has shipped yet.
  2. 2BI-FIND-2026-0178 · Six companies' agent-auth draft, since adopted by the WIMSE group, keeps credentials out of the model's reachEngineers from six companies, Defakto Security, AWS, Zscaler, Ping Identity, OpenAI and Okta, wrote a draft that opens by saying many agent authentication efforts "develop solutions in isolation, often reinventing existing mechanisms unaware of applicable prior art". It answers with a framework of existing standards and no new protocols: an agent is a workload with a WIMSE identifier that gets access through OAuth 2.0, and revocation and risk signals travel over the OpenID Shared Signals Framework. It still sets requirements of its own, such as the rule keeping the language model away from credentials, and says a framework like this helps find the gaps where extensions or profiles of existing standards are needed. The IETF datatracker records the WIMSE working group adopting the draft on 9 September 2026. It continues as draft-ietf-wimse-aims-00, posted on 15 September with the same six authors.
  3. 3BI-FIND-2026-0643 · A survey rates only workload identity ready for agents, and argues no standard can tell what an agent intendsIn the authors' grading of ten standards documents for agent identity, only the IETF's workload-identity work (WIMSE) with SPIFFE counts as available for production; SAML is graded incompatible, and the rest, including OAuth extensions for agents and the AI agent authentication draft whose security section still reads 'TODO', are partial or directional. The grades are the authors' own judgement over about 80 sources found by keyword search, with no scoring method or second rater, so they record what three researchers concluded in April 2026.
  4. 4BI-FIND-2026-0085 · Uber built the delegation chain the rest of the field calls unsolvedThe practitioner who built this says the standards are not ready to implement as written. Uber's exchange is 'conceptually based on' RFC 8693 and customized for its own auditing, and it names the IETF WIMSE working group and an individual draft on agent authentication as things it tracks for alignment rather than things it runs. The delegation-chain shape a large deployment needs is ahead of what the standards bodies have adopted.
  5. 5BI-FIND-2026-0020 · Give every AI agent a stable identifier the day it is created, then start with one team and one use caseAgent identifier standards are fragmented: schemes based on JSON Web Tokens (JWTs), the workload-identity standards SPIFFE and WIMSE, and proprietary formats, with no convergence across providers. Practitioners name interoperability as the gating factor and compare the moment to the early divergence of multi-cloud.
  6. 6BI-FIND-2026-0298 · The IETF's workload-identity architecture gives AI agents one subsection, calling an agent a delegated workloadThe IETF's workload-identity architecture does not give AI agents an identity type of their own. Its one passage on them says 'AI intermediaries are a special case of delegated workloads': an agent is software that inherits the security context of whoever called it and SHOULD pass that context on unless it has been authorized to reduce it. The one line it draws is between acting for someone and acting alone. Actions an agent takes that cannot be attributed to a principal MUST be clearly distinguished from delegated ones, for example with a separate workload identity or different token scopes. Vendors argue over whether agents are a third kind of identity; the standards text treats them as workloads with a caller.
  7. 7BI-FIND-2026-0299 · The IETF's workload-identity architecture gives AI agents one subsection, calling an agent a delegated workloadFor agents calling agents, the draft states the rule the rest of the field has been circling: 'each hop in the chain MUST explicitly scope and re-bind the security context', because otherwise a chain can extend authority well beyond what was granted. That is stronger than the OAuth token exchange standard, which leaves narrowing to the server's discretion. It is also only a requirement on paper. The document is informational, and the two protections it names, cryptographic binding of delegation tokens and attestation of what an intermediary did, are mentioned and not specified.
  8. 8BI-FIND-2026-0173 · OAuth's transaction token goes unchanged down a service chain and names who started it, not who acted in betweenThe draft keeps context and credential apart on purpose. A transaction token 'MUST NOT' be used by a workload to authenticate itself, 'MUST NOT' be used as an OAuth access token, must never carry an access token inside it, travels in its own header because the Authorization header 'may be used by the workloads for other purposes', and must not be logged verbatim because its request context can hold personal data such as the caller's IP address. Workload authentication is left to the WIMSE mechanisms. That is the opposite bet from the agentic JWT proposal, which binds an agent's intent into the access token itself. If the OAuth group's design holds, an agent's task description rides beside the credential in a second token, and any product that stuffs intent into the bearer token is off the standards track.
  9. 9BI-FIND-2026-0378 · The IETF's simplest workload proof proves the key and leaves the request unprotected, and the draft says so itselfA Workload Proof Token proves that the caller holds the key in its credential, and that is the limit of what it proves. The signed JWT carries the target URL as audience (query and fragment stripped), a hash of the workload's identity token, and hashes of any transaction token or other token in the request; it does not cover the HTTP method, the path beyond the audience, headers or the body. Section 3.3 states the consequence without softening it: the proof 'does not protect major portions of the request and response and therefore does not provide protection from an actively malicious middle box'. A workload that chooses this proof over the HTTP message-signature profile is trusting every proxy between it and the recipient with the content of the call.
  10. 10BI-FIND-2026-0377 · The IETF's workload credential binds one key to one name and says nothing else about the software holding itThe draft defines no way for a verifier to find out who to trust. Trust anchors belong to a trust domain and 'MUST be distributed out of band'; section 3 says the issuer claim 'MUST NOT be used to look up trust anchor material directly from information carried only in the token', which closes the key-fetch-by-URL shortcut OpenID Connect relies on. Federation between trust domains is not mentioned. Revocation is not solved either: section 9.2.1 states that JWT credentials 'are not generally revocable on demand', so a short expiry is 'the primary mitigation' for a stolen token and key. Sections 3 and 4 are new in this revision, so the working group added the out-of-band rule on purpose rather than leaving it unsaid.

›In our briefings

The claim each of the newest briefings on WIMSE would have you leave with.

  1. Uber runs per-hop token exchange with a full actor chain for AI agents in production, and has published sample tokens and a latency chart. Its security token service gives each hop a fresh token scoped to one audience, with a lifetime "in the order of minutes" (both sample payloads expire 300 seconds after issue), and the token carries the attested chain from the engineer to the agent making the call. Uber says "the system has been adopted by thousands of internal agents" and reports the exchange API's P99 as "consistently below 40 milliseconds" under current load; its chart sits mostly under 10 ms, with one spike reaching 40. The post covers this identity layer. Uber frames dynamic access control and a unified policy enforcement plane as its direction on top of it, and names session and intent claims as additions the token could carry in the future.

    Uber runs per-hop token exchange with a full actor chain for AI agents in productionBI-FIND-2026-0081

  2. Vajragowni's paper puts fixed numbers on governing non-human identities (NHIs). His lifecycle algorithm treats a credential older than 90 days as stale, rotating API keys and service-account passwords and renewing certificates. It calls an identity unused for more than 180 days dormant and quarantines any NHI with no owner at once, with a ticket to assign an owner or decommission it within 14 days. Identities both orphaned and unused for more than 365 days are decommissioned. His KPI table recommends targets such as under 2% of NHIs with no assigned owner or application, over 98% of credentials rotated within the policy window, and under 4 hours on average from a decommission trigger to revocation. The paper reports no deployment or test of these values and gives no reason for choosing them. Its own dormancy KPI counts identities with no activity in 90 days, where the algorithm waits 180.

    A paper's 90- and 180-day rules for machine identities come with no test behind themBI-FIND-2026-0030

  3. Revision -02 of Shankarlingam's draft says the delay before a disabled agent is refused is "a property of the receiver's configuration, not of the token alone", and that passing the receiver's checks does not show a disabled client is refused within the bound the receiver computes. Raising the maximum token age, DNS caching, caching of issuer metadata and keys, or the clock-skew allowance lengthens the time a disabled client is still accepted, while every signature, issuer, client, audience and DNS check keeps passing. Those checks, the draft says, "establish that a token is authentic; they do not establish that the client is still enabled." A receiver SHOULD recompute the bound before any change to those settings takes effect, and SHOULD test it directly, again after any change to those settings: obtain link tokens for a test client, disable it at a time not known to whoever observes the receiver, keep presenting tokens issued before the disable, and compare the last acceptance with the time the authorization server records the disable. The receiver meets its bound if that gap does not exceed the computed delay. It SHOULD also log the maximum age in effect with each verification result, and the security section says receivers SHOULD treat such changes as security relevant. The draft describes this test and reports no results from running it.

    A disabled agent's old tokens still pass every check; a DNS draft now asks receivers to time how longBI-FIND-2026-2345

›Cited in 82 briefings

Every briefing that discusses this term, newest first.

  1. Uber runs per-hop token exchange with a full actor chain for AI agents in production

    Article · · Uber Engineering blog · Matt Mathew et al. · May 21, 2026

  2. A paper's 90- and 180-day rules for machine identities come with no test behind them

    Article · · International Journal For Multidisciplinary Research (Vol 8, Issue 3) · Thrinaath Vajragowni · May 23, 2026

  3. A disabled agent's old tokens still pass every check; a DNS draft now asks receivers to time how long

    Draft · · IETF Datatracker · Sriram Shankarlingam, HexaEight · Oct 20, 2026

  4. AAuth gives every agent its own key, then keeps the delegation chain out of the token on purpose

    Draft · · IETF Datatracker · Dick Hardt, Hellō · Sep 25, 2026

  5. Asor's agent-chain draft now says a chain keeps verifying after its principal leaves, until expiry or revocation

    Draft · · IETF Datatracker · Rafael Asor · Oct 28, 2026

  6. Transaction Tokens carry who started a request through internal services and define no list of the hops in between

    Draft · · IETF OAuth Working Group · Atul Tulshibagwale et al. · Jul 30, 2026

  7. An agent-mandate draft checks who may delegate, and now intends to carry its chain in Asor's format

    Draft · · IETF Datatracker · Claudiu Hulea, Braincap SRL · Oct 24, 2026

  8. Agent-audit draft now admits a call on the principal and the current agent, as RFC 8693 requires

    Draft · · IETF Datatracker · Shameer Thaha et al. · Oct 19, 2026

  9. An agent-evidence draft no longer checks every actor in a delegation chain before a service admits a call

    Draft · · IETF Datatracker · Shameer Thaha et al. · Oct 19, 2026

  10. Two WIMSE participants say AI agents are workloads, and existing standards already cover most of the job

    Video · · Identerati Office Hours · Nov 6, 2026

  11. A Nokia and China Mobile draft finds A2A and MCP stop at one hop, and wants cancellation carried down the chain

    Draft · · IETF Datatracker · Tirumaleswar Reddy.K et al. · Oct 27, 2026

  12. A draft joining four agent records on one hash reports no valid joint run and now claims less about what agents did

    Draft · · IETF Datatracker · Steven Mih et al. · Aug 16, 2026

  13. SPIFFE trust domain names have no registrar, and the core spec warns that roles in a credential can go stale

    Article · · SPIFFE project (GitHub) · SPIFFE project · Jul 1, 2026

  14. Register a platform's identity provider once, not each agent: trust rests on its issuer, with no revocation

    Draft · · IETF Datatracker · Paul Carleton et al. · Sep 22, 2026

  15. WIMSE's agent draft adds how an agent's access ends and where policy engines fit, without a new requirement

    Draft · · IETF WIMSE Working Group · Pieter Kasselman et al. · Sep 15, 2026

  16. OAuth's SPIFFE draft opens metadata URLs to all three credentials and says what a self-served key cannot prove

    Draft · · IETF Datatracker · Arndt Schwenkschuster et al. · Jun 15, 2026

  17. Cisco's SD-Card rewrite pins MCP tools to what a registry signed; its security sections still read TODO

    Draft · · IETF Datatracker · Suhas Nandakumar et al. · Nov 1, 2026

  18. An agent audit draft says the agent is not trusted, yet lets the agent's operator run the audit store

    Draft · · IETF Datatracker · Mirja Kühlewind et al. · Sep 7, 2026

  19. An unadopted AGTP draft would let gateways refuse an agent's out-of-scope requests from its TLS certificate

    Draft · · IETF Datatracker · Chris Hood, Nomotic, Inc. · Oct 30, 2026

  20. Sweeney's delegation draft concedes its server is trusted with users' credentials, and drops an interop claim

    Draft · · IETF Datatracker · Kieran Sweeney · Oct 13, 2026

  21. An agent audit draft makes its first change under its format freeze, ahead of the tests for it

    Draft · · IETF Datatracker · Sankalp Gilda · Oct 16, 2026

  22. A 'human approved' field is not evidence, says Schrock's draft; only a bound, signed approval counts

    Draft · · IETF Datatracker · Iman Schrock, EMILIA Protocol · Jul 3, 2026

  23. WIMSE credentials -03: a valid signature proves trust-domain membership, not which workload is calling

    Draft · · IETF Datatracker · Brian Campbell et al. · Jul 2, 2026

  24. A Confidential Computing Consortium test for trustworthy workload identity rules out bearer tokens

    Article · · Confidential Computing Consortium (GitHub) · Trustworthy Workload Identity SIG et al. · Sep 13, 2026

  25. Agent URI draft -04 ties an agent's name to a key document and bars OAuth servers from trusting the name alone

    Draft · · IETF Datatracker · Yaswanth Narvaneni et al. · Oct 20, 2026

  26. Agent delegation records not anchored in trusted time become deniable once timestamps can be forged

    Draft · · IETF Datatracker · V.K. Uppalapati · Sep 29, 2026

  27. An agent audit log now names whom the agent acted for, but at low trust levels the agent may write that itself

    Draft · · IETF Datatracker · Raza Sharif, CyberSecAI · Oct 26, 2026

  28. Drake's revised draft: an agent identity outlives its operator but is never reissued once its last key is gone

    Draft · · IETF Datatracker · Christopher Drake, 1id.com · Sep 25, 2026

  29. Niyikiza's agent tokens narrow to a tool's arguments offline; revision -02 breaks -01 implementations

    Draft · · IETF Datatracker · Niki Aimable Niyikiza · Jun 15, 2026

  30. Attestation shows which image is running, not which copy; a new IETF draft weighs two ways to tell copies apart

    Draft · · IETF Datatracker · Tirumaleswar Reddy.K et al. · Oct 28, 2026

  31. A routing-policy draft requires authenticated links before hop checks can prove a chain

    Draft · · IETF Datatracker · Sumit P. Ahuja, Main Labs · Oct 21, 2026

  32. Delegation receipts for AI agents, revision 15: holding a copy of a receipt no longer lets you extend the chain

    Draft · · IETF Datatracker · Ryan Nelson, Authproof · Oct 22, 2026

  33. The IETF names workloads with a URI on the owner's domain; revision -04 pins down how two names are compared

    Draft · · IETF Datatracker · Yaroslav Rosomakho et al. · Oct 23, 2026

  34. An IETF draft says agent access checks may refuse when a source is down, but must not record that as a 'no'

    Draft · · IETF Datatracker · Girish Konda · Oct 19, 2026

  35. The IETF's stronger workload proof is built for the proxy in the middle, and it still leaves replay protection optional

    Draft · · IETF Datatracker · Joseph A. Salowey et al. · Oct 20, 2026

  36. An individual IETF draft takes a workload's role and group out of its name and puts them in separate metadata

    Draft · · IETF Datatracker · Shlomo Zalman Heigh et al. · Oct 16, 2026

  37. An 'approved' flag can record four different human acts, and only a check says anything about the thing approved

    Draft · · IETF Datatracker · Andreas Ehstand · Sep 27, 2026

  38. Six companies' agent-auth draft, since adopted by the WIMSE group, keeps credentials out of the model's reach

    Draft · · IETF Internet-Draft (individual submission, since adopted by the WIMSE working group) · Pieter Kasselman et al. · Jul 6, 2026

  39. Anthropic asks for per-agent identity; its guide leaves Claude Code sub-agent credentials unspecified

    Article · · Anthropic (Claude) · Anthropic · May 27, 2026

  40. A transparency-log profile for agents exposes edited and dropped records; an action nobody logged leaves no gap

    Draft · · IETF Datatracker · Pinar Emirdag et al. · Oct 12, 2026

  41. An end-to-end encrypted draft would hide an AI agent's task from the broker, but not who sent it or to whom

    Draft · · IETF Datatracker · Hamdy Atakora, Imara Labs · Sep 21, 2026

  42. Bu's checklist names a verifier for each agent-protocol claim and checks 'the same agent as before' separately

    Draft · · IETF Datatracker · Songbo Bu · Oct 14, 2026

  43. An attested OAuth client is one somebody vouched for; revision 12 says to trust its claims only as far as the voucher

    Draft · · IETF Datatracker · Tobias Looker et al. · Oct 13, 2026

  44. Schrock's receipts prove a named key approved one exact agent action, and the draft lists at length what they cannot prove

    Draft · · IETF Datatracker · Iman Schrock, EMILIA Protocol · Sep 12, 2026

  45. Munoz's agent Permit is signed before the action runs, but the profile does not make the action wait for it

    Draft · · IETF Datatracker · Christian Munoz, Keel API, Inc. · Jul 19, 2026

  46. The PEDIGREE draft has verifiers re-check every ancestor of an agent's token, without saying how to fetch them

    Draft · · IETF Datatracker · Karthik Rampalli, Glyphzero Labs · Apr 25, 2026

  47. OpenA2A's draft aims to keep secrets from the agent, and says no implementation yet supports its new controls

    Draft · · IETF Datatracker · Abdel Fane, OpenA2A · Oct 2, 2026

  48. Before an agent pays or changes data, Schrock's last check treats it as untrusted and keeps the record itself

    Draft · · IETF Datatracker · Iman Schrock, EMILIA Protocol · Sep 26, 2026

  49. Two implementations stop at the last of Reece's ten cross-company rules: a person approving one specific action

    Draft · · IETF Datatracker · Morgan Reece · Aug 31, 2026

  50. OpenA2A's protocol proves an agent holds a key and scores its behaviour, but never says who it works for

    Draft · · IETF Datatracker · Abdel Fane, OpenA2A · Oct 6, 2026

  51. Sharif's profile writes an AI agent's rights to pay and delete into an identity certificate valid up to a year

    Draft · · IETF Datatracker · Raza Sharif, CyberSecAI · Oct 2, 2026

  52. NIST's agent identity work starts from OAuth, SPIFFE and SCIM and defers agents from outside the enterprise

    Article · · NIST National Cybersecurity Center of Excellence · Harold Booth et al. · Feb 5, 2026

  53. AAuth's bootstrap guide makes device attestation optional and lets a new key mean a new agent

    Draft · · IETF Datatracker · Dick Hardt, Hellō · Sep 28, 2026

  54. A draft ties what each AI agent may read to what it may do, and still can't take a grant back

    Draft · · IETF Datatracker · Hamdy Atakora · Sep 21, 2026

  55. DIF's agent spec names every AI agent by a DID; the IETF's Agent Name Service switched to domain names

    Article · · Decentralized Identity Foundation (KYA-OS) · Dylan Hobbs et al. · Jul 29, 2026

  56. A survey rates only workload identity ready for agents, and argues no standard can tell what an agent intends

    Paper · · arXiv · Takumi Otsuka et al. · Apr 25, 2026

  57. An agent audit record that shows when a 'deny' went through anyway

    Draft · · IETF Datatracker · Sankalp Gilda · Oct 5, 2026

  58. Folding every hop's signature into one would stop anyone stripping a signed hop out of an agent's request

    Draft · · IETF Datatracker · Tirumaleswar Reddy.K et al. · Sep 28, 2026

  59. WIMSE mutual TLS -03 ties certificate acceptance to trust-domain anchors

    Draft · · IETF Datatracker · Joseph A. Salowey et al. · Jul 6, 2026

  60. Cisco's SD-Card puts an A2A Agent Card in a selective-disclosure token, and leaves its registry out of scope

    Draft · · IETF Datatracker · Suhas Nandakumar et al. · Mar 2, 2026

  61. A third SCIM agent draft gives the agent five lifecycle states, a quarantine, and an archive it can never leave

    Draft · · IETF Datatracker · Saurabh Kushwaha, Oracle · Jul 27, 2026

  62. The IETF's simplest workload proof proves the key and leaves the request unprotected, and the draft says so itself

    Draft · · IETF Datatracker · Brian Campbell et al. · Aug 27, 2026

  63. The IETF's workload-identity architecture gives AI agents one subsection, calling an agent a delegated workload

    Draft · · IETF Datatracker · Joseph Salowey et al. · Jul 6, 2026

  64. The 'intent' in the Agentic JWT draft is a workflow step, and the agent reports its own fingerprint

    Draft · · IETF Datatracker · Abhishek Goswami · Jul 1, 2026

  65. Highflame's open-source ZeroID narrows an agent's token at every hop, a step we did not find in Uber's write-up

    Article · · Highflame (GitHub) · Sep 25, 2026

  66. Agent identifiers are split across JWT-based, SPIFFE, WIMSE and proprietary schemes, an identity lead says

    Podcast · · Identity Decoded · Winning in securing agentic identity: Start with one team, one use case · Aug 12, 2026

  67. IPC passes a delegation chain between services; its signature proves only the last signer

    Draft · · IETF Datatracker · Ravi Kant Sharma, Ericsson · Jul 23, 2026

  68. A valid delegation chain cannot replace required human approval or proof of key possession

    Draft · · IETF Datatracker · Karthik Rampalli, Glyphzero Labs · Jul 6, 2026

  69. A June draft moves workload attestation out of TLS into HTTP headers so it survives a TLS-ending proxy

    Draft · · IETF Datatracker · Tirumaleswar Reddy.K et al. · Jun 7, 2026

  70. An agent certificate draft ties access to a trust score set by the issuing CA, usually the agent operator's own

    Draft · · IETF Datatracker · Raza Sharif, CyberSecAI · Apr 10, 2026

  71. An agent-credential draft carries a hash of the user's instruction that its own verification steps never check

    Draft · · IETF Datatracker · Chudah Yakung, Attest · Mar 26, 2026

  72. AIP puts an agent's delegation chain in the token; its perfect score against attacks tests rules only AIP has

    Paper · · arXiv · Sunil Prakash · Mar 25, 2026

  73. A Huawei agent draft keeps one 'master agent' as gatekeeper and drops the line that had servers re-check it

    Draft · · IETF Datatracker · Yuan Ni et al. · Feb 28, 2026

  74. An Alibaba and Cisco draft signs what the user was shown, then enforces a policy the agent wrote

    Draft · · IETF Datatracker · Dapeng Liu et al. · Feb 28, 2026

  75. A draft binds workload credentials to their host via a local agent, and leaves that agent's compromise undecided

    Draft · · IETF Datatracker · Ram Krishnan et al. · Feb 26, 2026

  76. A 2025 draft would map attestation results to the business terms that access policies use

    Draft · · IETF Datatracker · Mark Novak et al. · Nov 5, 2025

  77. Rosenberg and White's AAuth now has the resource server, not the agent, write and sign the approval text

    Draft · · IETF Datatracker · Jonathan Rosenberg et al. · Oct 20, 2025

  78. Attested workload identities need a mapping that survives authorized updates

    Draft · · IETF Datatracker · Mark Novak et al. · Oct 20, 2025

  79. WIMSE leaves its application proof choice open, with different integrity coverage

    Draft · · IETF Datatracker · Brian Campbell et al. · Oct 16, 2025

  80. The 2024 'IDs for AI' paper puts the ID on a chat session, not a model, and leaves trust in its issuer to others

    Paper · · arXiv · Alan Chan et al. · Jun 17, 2024

  81. Richer's 2023 draft let each service in a call chain add a hash-linked token, but left serialization unfinished

    Draft · · IETF (individual Internet-Draft) · Justin Richer · Sep 15, 2023

  82. A 2025 workload identity draft kept credential exchange for crossing trust domains, and never to gain more trust

    Draft · · IETF Datatracker · Arndt Schwenkschuster, Defakto · Mar 3, 2025

In our columns

  1. Switching off an AI agent needs a receipt from every system that held its access

    Column · Nov 7, 2026

  2. Prompt injection tests now hit 2%; the control that holds sits outside the model

    Column · Oct 18, 2026

  3. Gartner's identity curve needs a closer look at what has actually shipped

    Column · Oct 15, 2026

  4. Machine identities lack a shared schema, so buy discovery tools that export their data

    Column · Sep 17, 2026

  5. Agent identity's requirements are settled. Buy those, not a token format

    Column · Sep 16, 2026

Related terms

What appears alongside WIMSE in the same briefings.

SPIFFEWIMSE Workload Proof TokenWIMSE ArchitectureTransaction TokensOAuth Token ExchangeHTTP Message SignaturesRATSJWT

Was this explainer useful?

Share this explainer

Pass it on. Every claim has a permanent number, so what you share today still points at the same words next year.