OpenID Connect
OIDC
By Boston Identity · Explained · 6 min read
The short answer
OpenID Connect is a thin identity layer on top of OAuth 2.0. An identity provider signs the user in and gives the app a signed ID token saying who the user is and which app the token is for.
- What it is
- A protocol
- Defined by
- OpenID Foundation, Core 1.0 (final, errata set 2)
- Who it's for
- Teams running sign-in for apps, APIs and agents
- In our briefings
- 1380by month, last 12
What is OpenID Connect?
OAuth 2.0 lets an app get an access token to call an API on someone's behalf. By itself it has no standard way to tell the app who that someone is.§1 OpenID Connect (OIDC) fills that gap. The specification calls itself "a simple identity layer on top of the OAuth 2.0 protocol" that lets an app verify the user's identity from a sign-in done by the provider.§1
Two roles carry new names. The identity provider is the OpenID Provider (OP), and the app relying on its answer is the Relying Party (RP).§1 An app turns OIDC on by adding the openid scope to an ordinary OAuth request.§1 The answer comes back as an ID token, a JSON Web Token (JWT), meaning a compact signed JSON document.§2
The ID token must carry the issuer (iss), the subject identifier for the user (sub), the audience (aud), an expiry (exp) and an issue time (iat).§2 It may also say when the user signed in (auth_time) and which sign-in methods were used (amr).§2 It must be signed, and it may also be encrypted.§2 Core 1.0 is a final specification; the copy we read incorporates errata set 2, dated 15 December 2023.S1
Not the same as OAuth 2.0, which it is built on. OAuth defines how an app gets an access token for an API, and without a profile like this one it says nothing about how the user signed in.§1
How does OpenID Connect work?
Three parties take part: the user with a browser, the app (the Relying Party), and the identity provider (the OpenID Provider) with its authorization, token and UserInfo endpoints. Core defines three flows; this is the Authorization Code Flow, where every token comes back from the token endpoint.§3, §3.1
Figure steps
- The app sends the user to the provider
- The provider signs the user in
- The provider gets an authorization decision
- The provider sends the user back with a code
- The app trades the code for tokens
- The app validates the ID token
- The app may ask for more about the user
In the code flow the browser carries only a code; the app collects the ID token straight from the provider's token endpoint and checks it before trusting who signed in.
Drawn from OpenID Connect Core 1.0, sections 1.3 and 3.1
Swipe for the full figure →
The app sends the user to the provider
The app builds an authentication request and sends the browser to the provider's authorization endpoint over TLS (Transport Layer Security).§3.1.2 The request must contain the openid scope, the response type code, the app's client_id and a redirect_uri that exactly matches one the app registered.§3.1.2.1 A state value is recommended against cross-site request forgery, and a nonce, a random value later echoed in the ID token, is optional in this flow.§3.1.2.1
The provider signs the user in
The provider authenticates the user, or checks that they are already signed in.§3.1.2.3 How it does that, by password or session cookie for example, is outside the specification. With prompt=login it must sign the user in again; with prompt=none it must not interact with the user, and must return an error if the user is not already signed in.§3.1.2.3
The app trades the code for tokens
The app posts the code to the provider's token endpoint over TLS, and a confidential client, one that can keep a secret, must authenticate itself there.S1 The provider checks that the code is valid and was issued to this client, and that the redirect_uri matches the original request.§3.1.3.2 It answers with an ID token and an access token.§3.1.3.3
The app validates the ID token
The app must check that iss exactly matches the provider's issuer, that aud contains its own client_id, and that the token has not expired.§3.1.3.7 It must reject a token that lists other audiences it does not trust.§3.1.3.7 It must check the signature with the issuer's keys, though in this flow it may rely on the TLS connection to the token endpoint instead.§3.1.3.7 If it sent a nonce, the same nonce must be in the token.§3.1.3.7 The app then takes the user's subject identifier from the token.§3.1.1
The app may ask for more about the user
Optionally, the app sends the access token to the provider's UserInfo endpoint, which returns claims about the user such as name or email.§1.3, §5.1 The specification defines the UserInfo endpoint as a protected resource that must use https.§1.2
Why does OpenID Connect matter now?
Vendors are still changing how they issue OIDC tokens this year. Okta made encryption of OIDC ID tokens generally available for custom app integrations, and Ping said that from 2 March 2027 PingOne will issue only opaque refresh tokens for OIDC apps.1 Okta also made client ID metadata documents generally available in preview for OIDC apps on 16 September.2
OIDC has moved past the sign-in button, too. GitHub extended npm trusted publishing over OIDC, and AWS took outbound OIDC federation onto private networks.3 After a worm spread on stolen developer credentials, Pandya named OIDC-based trusted publishing as the replacement for long-lived npm tokens.4
Agents are the open question. NIST's National Cybersecurity Center of Excellence lists OpenID Connect among the standards it is "currently considering" for agent identity.5 Core has no claim saying what kind of subject sub names, and no claim for one party acting for another.6 A draft OpenID profile adds a sub_type claim marking the subject as a user, an application or an agent.7
Where does it fit?
| Compared with | How OpenID Connect differs |
|---|---|
| OAuth | OAuth 2.0 is the base: it gets an app an access token for an API. OpenID Connect adds the openid scope and the ID token so the app also learns who signed in and how.§1 |
| SAML | Small says most big clouds support both SAML and OpenID Connect, and that this alone is not enough to leave a cloud.8 OpenID Connect carries its sign-in result as a signed JWT on top of OAuth 2.0.§1, §2 |
| JWT | A JWT is a token format. An ID token is one particular JWT, with required claims for issuer, subject, audience, expiry and issue time.§2 |
| CIBA | CIBA is an OpenID flow in which the person approves on another device. Its provider must still obtain the person's decision as OpenID Connect Core describes it.9 |
What goes wrong?
- Trusting the wrong identifier. Only iss and sub together identify a user; email, phone number and username must not be used as unique identifiers, because a provider may give one email address to different people over time.10
- Skipping validation. An app that does not check the issuer, audience, expiry and nonce can accept a token meant for someone else.§3.1.3.7 Core forbids the unsigned alg value none except in narrow cases the client asked for at registration.§2
- Choosing the Implicit Flow without weighing its limits. There the tokens come straight back through the browser, where the user and other apps with access to it can see them, and the client cannot be authenticated.§3.2 Core says the flow may be appropriate for a public client but never mentions PKCE (Proof Key for Code Exchange), and OAuth's later security guidance tells clients to avoid the implicit grant, with an exception.11
- Mixing up providers. A client that works with more than one authorization server is required to defend against mix-up attacks, for example by checking the ID token's iss claim.12
- Assuming logout is covered. Core does not say how a sign-in ends; logout lives in separate optional specifications.13
What should a security leader ask?
- 01
Which of our apps match users by email address instead of by issuer and subject together?
- 02
Which of our apps still use the Implicit Flow, and which ones skip the nonce or ID token checks?
- 03
When someone is disabled at our identity provider, how does each app learn to end that session?
- 04
Which apps accept ID tokens from more than one provider, and do they check the issuer on every response?
- 05
When one of our agents signs in, what in its token tells a service that it is an agent and not a person?
Sources
- S1
OpenID Connect Core 1.0 incorporating errata set 2OpenID Foundation · Final, errata set 2 (15 December 2023)
›Evidence: 13 claims from 9 briefings
- 1BI-FIND-2026-1173 · In January Okta's agent product was in early access; Auth0 and Microsoft made machine-to-machine features GAOkta and Ping both changed token formats in January, and Ping's change comes with a deadline. Okta made JSON Web Encryption of OIDC ID tokens generally available for custom app integrations. PingOne began offering opaque refresh tokens for OIDC apps on January 27, and its note says that from March 2, 2027 PingOne will issue only opaque refresh tokens; apps not moved off JWT refresh tokens by March 1, 2027 may leave users unable to reach resources. We read both as limiting who can read a token's contents to the party meant to read it.
- 2BI-FIND-2026-0704 · Three vendors now route MCP tool calls through an identity check, and only Cloudflare's is generally availableOkta and Ping both shipped client ID metadata documents (CIMD), and CyberArk went the other way. With CIMD, an agent or app names itself with a URL that points to its own metadata, and needs no prior registration. Okta made CIMD a beta for AI agents and GA in preview for OIDC apps on September 16. Ping's Advanced Identity Cloud shipped it on September 29. CyberArk's September MCP feature relies on dynamic client registration (DCR) instead. Okta and Ping also changed their DCR support in the same releases, so neither has dropped it.
- 3BI-FIND-2026-0707 · Three vendors now route MCP tool calls through an identity check, and only Cloudflare's is generally availableMost of September's non-human identity releases find and list credentials, and fewer remove them. SailPoint added thirteen connectors that list tokens, keys and service accounts. GitHub made credential inventory exports GA, and Cloudflare now cleans up inactive service tokens. Releases that replace stored secrets with a workload identity were fewer: CyberArk added workload attestation for AWS and Kubernetes, GitHub extended npm trusted publishing over OIDC, and AWS took outbound OIDC federation onto private networks.
- 4BI-FIND-2026-1533 · Shai-Hulud v4 used an MCP tool description to tell AI assistants to read SSH keys and keep quiet about itThe worm spreads on stolen developer credentials, tried one after another. With an npm publish token it first checks the token works and skips accounts protected by two-factor authentication, which Pandya says "cannot be quietly abused". It ranks the remaining packages by how widely they are used and how abandoned they look, then publishes a patch-version bump after a random delay. With a GitHub token it commits a carrier dependency and a CI workflow, and its last fallback is the SSH agent. To persist, it points the global git init template directory at infected hooks. For hardening he names OIDC-based trusted publishing in place of long-lived npm tokens, pinned GitHub Actions, and review of every workflow and dependency change.
- 5BI-FIND-2026-0686 · NIST's agent identity work starts from OAuth, SPIFFE and SCIM and defers agents from outside the enterpriseNIST's National Cybersecurity Center of Excellence framed agent identity in February as a question of applying what exists, not inventing something new. Its concept paper lists OAuth, OpenID Connect, SPIFFE, SCIM, the NGAC access-control standard and MCP as candidates it is "currently considering", and names none of the agent-specific IETF work, such as ID-JAG, the WIMSE drafts or token exchange for agents. It keeps human delegation and logging in scope and leaves agents from outside the enterprise for a later effort. The next steps are a possible project description and a lab-built practice guide, with no date; the comment period closed on 2 April.
- 6BI-FIND-2023-0057 · OpenID Connect Core says only issuer plus subject identify a user; email must never be the keyCore writes the ID token as a statement about one person's sign-in, addressed to the client that asked for it. It defines the End-User as a "Human participant". The aud claim MUST contain the relying party's client_id and MAY name other audiences, and the client MUST reject a token that does not list it or lists audiences it does not trust. The azp claim, which names the party the token was issued to, is optional; the spec notes that "in practice, the azp Claim only occurs when extensions beyond the scope of this specification are used". Core's text defines no claim saying what kind of subject sub names and none for one party acting for another; the word agent appears only in "User Agent". We read that as the gap agent profiles fill: our briefing on Raza Sharif's OpenID agent-claims draft found it adding a sub_type claim that labels the subject a user, an application or an agent.
- 7BI-FIND-2026-1397 · OpenID agent-claims draft adds a provider-set user, application or agent label; scores stay each provider's ownRevision -02 adds an eleventh claim, sub_type, that says whether the token's subject is a user, an application or an agent. The identity provider sets it from its own registration record, and it MUST NOT be suppliable or overridable by the client. Relying parties now MUST NOT infer an agent from agent_ claims, which reverses -00's line that the agent_ prefix lets them identify agent-specific assertions unambiguously. They MUST validate the issuer before relying on sub_type and MUST treat a missing value as unknown, failing closed where an authorization decision depends on the class. The claim sits outside the draft's own checks: none of the ten validation steps relying parties MUST run in Section 7.1 tests it. Section 5 requires it in an Agent ID Token, yet a provider MAY let an administrator omit it for a given client. When an agent acts for a user, the top-level sub_type describes the user and the act claim SHOULD carry the agent's own. The draft does not say whether such a token counts as an Agent ID Token, where sub_type is REQUIRED and the Section 5 example carries "agent". Sharif credits the design to three reviewers from WSO2's ThunderID project and says ThunderID shipped it in August 2026, but the draft links no release.
- 8BI-FIND-2026-1401 · KuppingerCole's Mike Small blames most cloud incidents on customers' identity mistakes and sees CNAPP moving to cover AISmall says most big clouds support SAML and OIDC, but that is not enough to leave one: moving clouds means moving every identity permission, so he suggests asking a provider whether those permissions can be exported. Vendors help customers in, he says, but not out, and standards help without being sufficient.
- 9BI-FIND-2021-0001 · CIBA's human-approval step carries a short matching code, not a description of the actionIn CIBA, the one request parameter meant to be shown to the person is binding_message, and the spec designs it as a short code for matching two screens. It says the value SHOULD be relatively short, in a limited set of plain-text characters, and should let the user tell that the two devices belong to one transaction, "such as a random value of reasonable entropy (e.g. a transactional approval code)." Past that, the provider MUST obtain the person's authorization decision as OpenID Connect Core describes it. The spec defines no parameter for an action or an amount, though it lets extensions and profiles add parameters, and none of the tokens it defines carries the binding message back to the client. Several agent designs we have read cite CIBA for approving a specific action; we read that as needing a profile on top of the core spec.
- 10BI-FIND-2023-0054 · OpenID Connect Core says only issuer plus subject identify a user; email must never be the keyOpenID Connect Core gives an app one identifier it can rely on for a user: the iss and sub claims together. The sub value MUST be locally unique and never reassigned within the issuer, and it MUST NOT exceed 255 ASCII characters. Email, phone_number, preferred_username and name MUST NOT be used as unique identifiers, because an issuer may give one email address to different people over time and a person's address may change. When the app also calls the UserInfo endpoint, the sub returned there MUST exactly match the ID token's, or the response values MUST NOT be used.
- 11BI-FIND-2023-0060 · OpenID Connect Core says only issuer plus subject identify a user; email must never be the keyCore's advice for browser and native apps sits apart from OAuth's later security guidance. Section 15.4 says the Implicit Flow "may be appropriate" for a public client, such as an app running in the browser or a statically registered native app. In that flow every token comes back from the authorization endpoint and the client cannot be authenticated; Core requires a nonce in that flow and defines an at_hash claim that ties a returned access token to the ID token. Core's text never mentions PKCE. Our briefing on RFC 9700, OAuth's security best current practice, found it telling clients to avoid the implicit grant, with an exception, and requiring public clients to use PKCE.
- 12BI-FIND-2025-0353 · RFC 9700 bans OAuth's password grant and loose redirect matching; binding access tokens stays a SHOULDA client that can work with more than one authorization server is REQUIRED to defend against mix-up attacks. The attack assumes one of those servers is under the attacker's control, which the RFC says can happen, for example, when the attacker registers the client at their own server through dynamic registration, or when a server is compromised. In its walk-through the user picks the attacker's server, the attacker forwards the request to the honest server under the client's ID there, and the client, still believing the attacker's server answered, redeems the honest server's code at the attacker's token endpoint. Clients should check the iss parameter (RFC 9207) or an issuer value such as the ID Token's iss claim. Distinct redirect URIs per server also work, but the RFC says that defence "SHOULD therefore only be used if other options are not available".
- 13BI-FIND-2023-0059 · OpenID Connect Core says only issuer plus subject identify a user; email must never be the keyCore does not say how a sign-in ends. Its text mentions logout only in its references and in section 15.7's list of related optional specifications: RP-Initiated Logout, Session Management, Front-Channel Logout and Back-Channel Logout. Inside Core, the related rules are about token lifetime. Access tokens "might not be revocable by the Authorization Server", so their lifetimes SHOULD be kept to single use or very short. The provider MUST always obtain the user's consent before returning a refresh token for offline access, and an ID token issued on refresh keeps the original iss, sub and aud and, if present, the original auth_time.
›In our briefings
The claim each of the newest briefings on OpenID Connect would have you leave with.
OMB's post-quantum memo M-26-15 says "API gateways and application workloads must be configured to issue and validate PQC-signed tokens", naming JWTs as an example, and gives that line no date or phase of its own. The same appendix says cryptographic device attestation, often rooted in a Trusted Platform Module, must be migrated to post-quantum algorithms. In the body, agencies should count logical access control systems based on asymmetric encryption, such as public key infrastructure, that use any of the algorithms listed in Appendix A among the systems they prioritize. GSA will set up an inter-agency working group within 60 days on modernizing Federal Identity, Credential, and Access Management (FICAM) for post-quantum cryptography, and GSA's FICAM office is working with interested agencies to test PQC-ready physical and logical access systems. The memo's text, appendices included, never mentions AI agents, machine identities, OAuth or code signing.
OMB tells agencies their API gateways must issue post-quantum signed tokensBI-FIND-2026-0687
Auth0 Agent Gateway, badged 'Beta Now' on its slide, is one hub between an organisation's agents and the MCP servers they call. Auth0 says it can handle access policies, credential injection, audit logging and organisation boundaries, and the slide adds a kill switch. The console in the demo already lists Asana, Atlassian, BambooHR, GitHub, Notion, Salesforce, Sentry, Slack and Workday as servers, counts '31 of 200 tools max', and labels MCP Servers beta and a separate Agents section preview. One HR agent, with the same code and one deployment, reaches Workday and Slack for one customer tenant and BambooHR and Slack for another, because the gateway handles the token exchanges and routes the agent to the right tool for each customer. Our briefing on WorkOS's account of Sierra's internal gateway found that access control and audit held up that rollout, and we read Agent Gateway as a product for that layer. The keynote does not mention Ping Identity, but our briefing on Ping's March 2026 launch video covers a Ping product also called Agent Gateway.
Auth0 ships an agent gateway in beta and moves the identity form into the chat so the agent never sees the cardBI-FIND-2026-0214
Linx's Goldenberg calls one agent acting for one user "in a way" a solved challenge, because the user's token usually already carries what downstream applications need to control data access. Once agents delegate to other agents across several hops, he says, "you lose track of who is the initial actor", and he sees toxic combinations and violations of segregation of duties. Aembit's Goldschlag answers with token exchange at every hop, from user to agent to gateway to MCP server to resource, each hop holding only the credential for the next step. Godha points to IETF drafts on chained delegation tokens and to transaction tokens, which he says are coming.
When AI agents hand work to other agents, the person who started the job can get lostBI-FIND-2026-0070
›Cited in 138 briefings
Every briefing that discusses this term, newest first.
- OMB tells agencies their API gateways must issue post-quantum signed tokens
Article · · Office of Management and Budget (whitehouse.gov) · Russell T. Vought et al. · Jun 24, 2026
- Auth0 ships an agent gateway in beta and moves the identity form into the chat so the agent never sees the card
Video · · Okta · Uploaded September 24, 2026
- When AI agents hand work to other agents, the person who started the job can get lost
Podcast · · The Non-Human & AI Identity Podcast · NHI & AI Identity Summit : Governing AI · Sep 7, 2026
- MCP's July spec ties client credentials to the authorization server that issued them
Article · · Model Context Protocol project · Jul 28, 2026
- 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
- 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
- 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
- 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
- Sello has the services an AI agent calls sign its receipts; the agent can't fake them, but it can go around them
Paper · · arXiv · Juan Figuera · Jun 2, 2026
- MCP's client credentials extension puts the secret in the form body, while its server rule names HTTP Basic
Article · · Model Context Protocol project · Oct 23, 2026
- 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
- 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
- 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
- OAuth's draft for token requests that wait on a person now gives the client a link for the user to approve
Draft · · IETF Datatracker · Guilherme de Oliveira Niero et al. · Sep 23, 2026
- Cross-App Access draft warns that one stolen sign-in token could reach every app the identity provider permits
Draft · · IETF Datatracker · Aaron Parecki et al. · May 21, 2026
- Give an agent access for the task and take it back afterwards, four voices say, and four say people approving each request cannot keep up
Field signals · · 8 voices from the field · Feb 2026 to Oct 2026
- Global Token Revocation must end a user's refresh tokens; their access tokens can outlive the call
Draft · · IETF Datatracker · Aaron Parecki, Okta · Feb 25, 2026
- AGTP checks that an agent ID is signed; binding it to the connection is an add-on in the author's patent notice
Draft · · IETF Datatracker · Chris Hood, Nomotic, Inc. · Jun 28, 2026
- NovaFabric seals AI agent runs as signed evidence, and its author says a signature proves a key, not a person
Paper · · arXiv · Mohsen Seyedkazemi Ardebili · Oct 25, 2026
- Receiver acknowledged, gateways cut off: in a fixed 16-host replay, 112 of 225 revoked tool calls still committed
Paper · · arXiv · Lifei Liu et al. · Oct 11, 2026
- MCP's November 2025 authorization spec puts metadata documents ahead of dynamic client registration
Article · · Model Context Protocol · Nov 25, 2025
- 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
- A cross-domain transaction-token grant now names AI agents as a use and keeps the call chain at home
Draft · · IETF Datatracker · George Fletcher et al. · Sep 10, 2026
- Shai-Hulud v4 used an MCP tool description to tell AI assistants to read SSH keys and keep quiet about it
Video · · OWASP Foundation · Uploaded October 22, 2026
- Five places to put authorization in microservices, and why Drus says the classic gateway one does not scale
Video · · OWASP Foundation · Uploaded October 21, 2026
- IBM Research puts agent identity in a SPIFFE and OAuth sidecar; Maia Iyer calls authorization policy still open
Podcast · · Kubernetes Bytes · Your AI Agent Has Too Much Access | Fixing Agent Identity · Sep 15, 2026
- AAuth carries the signals an agent's approval needs, Hardt says; the brain that decides does not exist yet
Podcast · · Collateral Dialogue · Self-Driving Software: What OAuth Can’t Do for AI Agents | Dick Hardt, Hello · Sep 2, 2026
- RFC 9943 makes a log prove who signed a statement, not that it is true or that nothing was left out
Article · · IETF (RFC Editor) · Henk Birkholz et al. · Jun 30, 2026
- Best Buy shrank its identity stack to three tools in 2021 and now throttles AI agents ahead of an RFP
Podcast · · Identity at the Center · #411 - Making IAM a Best Buy with Greg Handrick · Mar 30, 2026
- KuppingerCole's Mike Small blames most cloud incidents on customers' identity mistakes and sees CNAPP moving to cover AI
Podcast · · Identity at the Center · #455 - CNAPP and the Cloud Identity Problem with Mike Small · Oct 19, 2026
- ChainIT's design has the payment rail rebuild each payment's hash before money moves; the paper reports no test
Paper · · arXiv · Jeremy Blackburn et al. · Oct 16, 2026
- 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
- An OAuth profile for AI agents makes every hand-off cut privileges, but a shorter expiry alone counts as a cut
Draft · · IETF Datatracker · Angel Cruz · Feb 7, 2026
- OpenID Federation 1.0 lets each level of a trust chain only tighten the rules, and defines no revocation process
Article · · OpenID Foundation · Roland Hedberg et al. · Feb 17, 2026
- Okta's expired SCIM agent draft recorded certificates and OAuth clients; the joint draft after it dropped both
Draft · · IETF Datatracker · Macy Abbey et al. · Oct 16, 2025
- A decentralized ID design for agents proves who is liable but can collect only from collateral locked in advance
Paper · · arXiv · Ahammod Munim Ramil et al. · Oct 12, 2026
- 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
- 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
- A2A 1.0 connects agents across companies and leaves out the person they act for
Article · · A2A Protocol (Linux Foundation) · A2A Protocol project et al. · May 28, 2026
- 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
- Auth0's own Claude Code tutorial runs the agent on the developer's login and hands it fifteen admin scopes
Article · · Auth0 Blog · Manish Hatwalne · Sep 29, 2026
- Three vendors now route MCP tool calls through an identity check, and only Cloudflare's is generally available
Release tracker · · 11 vendors' release notes · Sep 2026
- OWASP's agent Top 10 blames identity systems built for people, and does not say how it chose or ranked its risks
Article · · OWASP GenAI Security Project · John Sotiropoulos et al. · Dec 9, 2025
- Okta says its agent kill switch revokes all tokens; its product screen says tokens outside the gateway survive
Video · · Okta · Sep 24, 2026
- 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
- NIST and CISA advise one-hour tokens, 90-day keys and hardware key storage; AI agents get one paragraph
Article · · NIST (CSRC) · Ryan Galluzzo et al. · Sep 15, 2026
- MCP's enterprise login extension says IdP revocation is immediate. Its own spec does not back that up
Article · · Model Context Protocol (modelcontextprotocol.io) · MCP project et al. · Jun 18, 2026
- 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
- 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
- An integrator builds agent access on OpenID Connect and OAuth; clients want one view across many MCP gateways
Podcast · · Identity Unfiltered · CIAM, Agentic AI & the Future of Identity · Oct 5, 2026
- Transaction tokens are nearly a standard, and there is still no agreed place in them for the agent
Podcast · · Identity at the Center · #449 - Decoded - Transaction Tokens with George Fletcher · Sep 21, 2026
- WorkOS's Tobin South, in November 2025: agents will get a new kind of account, not a new identity standard
Podcast · · CodingCat.dev Podcast · 5.10 - How to Secure AI - Identity Access Management for Agents · Nov 21, 2025
- One shared service account turned a one-day password change into three weeks
Podcast · · The Non-Human & AI Identity Podcast · Securing NHI and Human Identity and Rise of Agentic AI · Jan 7, 2026
- MCP's authorization spec makes the agent a plain OAuth client and leaves delegation to an extension
Article · · Model Context Protocol project · Jul 28, 2026
- 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
- The agent risk two CISOs fear is manipulation, not a stolen credential
Podcast · · Defense in Depth · Identity and Access Management (IAM) in an Agentic AI World · Jul 23, 2026
- Ping moved agent identity from roadmap to shipping product, and let AI agents run its admin tools too
Conference · · Ping Identity · Austin · 2-day event (Day 1 / Day 2), 2026
- IPC passes a delegation chain between services; its signature proves only the last signer
Draft · · IETF Datatracker · Ravi Kant Sharma, Ericsson · Jul 23, 2026
- xChk ships human OIDC portfolios and a separate agent approval API
Paper · · arXiv · Sean MacGuire · Jul 15, 2026
- The browser OAuth draft favors a backend that keeps tokens out of JavaScript
Draft · · IETF Datatracker · Aaron Parecki et al. · Jul 6, 2026
- One implementer found a transaction-token rule that let any scope through; the OAuth group fixed it in weeks
Draft · · IETF Datatracker · Dhruv Agnihotri · Jun 23, 2026
- GM’s employee directory gave applications a reason to adopt single sign-on
Podcast · · Identity Decoded | The Identity Security Podcast · From IT support to security's core: General Motors' identity story · Jun 17, 2026
- OpenID's AuthZEN group approves a draft that lets policy tell an agent 'not yet, and here is what is required'
Article · · OpenID Foundation · Serj Hallam · Jun 15, 2026
- Atlassian's first draft for offline delegation banned bearer handles; revision 05 just recommends sender binding
Draft · · IETF Datatracker · Larry Zhu et al. · May 22, 2026
- 114 of 119 MCP servers that let outsiders register an OAuth client let them choose where sign-in codes are sent
Paper · · arXiv · Huijun Zhou et al. · May 21, 2026
- A proposed consent log for A2A and MCP records the terms an agent accepted, but the agent fills it in itself
Paper · · arXiv · Ravi Kiran Kadaboina · Apr 16, 2026
- Microsoft and Google made AI agent identity generally available in April; Okta's notes put its GA in May
Release tracker · · 8 vendors' release notes · Apr 2026
- OpenID agent-claims draft adds a provider-set user, application or agent label; scores stay each provider's own
Draft · · IETF Datatracker · Raza Sharif, CyberSecAI · Mar 26, 2026
- 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
- Outside Azure, Argo CD uses stored secrets for repositories; a March plan for per-project identity is still open
Video · · CNCF [Cloud Native Computing Foundation] · Uploaded March 24, 2026
- An agent identity draft signs every tool call, but never finished saying whom the agent answers to
Draft · · IETF Datatracker · James Cao et al. · Mar 16, 2026
- Auth0's guide ties sessions to TLS fingerprints, including JA3, which Chrome changes on each new connection
Article · · Auth0 Blog · Antonio Ortells · Mar 5, 2026
- OAuth 2.1's October draft bans plain PKCE and adds browser-app rules; 2.1-only clients can fail on older servers
Draft · · IETF Datatracker · Dick Hardt et al. · Mar 2, 2026
- An OAuth client brokering many tools, as agent services do, can let a rogue tool take over a user's other tools
Draft · · IETF Datatracker · Tim Würtele et al. · Mar 2, 2026
- 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
- Wisconsin researchers check each agent action against its history, in small tests with hand-written rules
Paper · · arXiv · Nils Palumbo et al. · Feb 18, 2026
- Auth0's AgentCore guide says prompts alone are not enough, then lets the prompt decide when to ask a person
Article · · Auth0 Blog · Kapil Patil · Feb 10, 2026
- An open spec for stopping agent actions before they run needs the vendor's help to cover ChatGPT or Copilot
Paper · · arXiv · Herman Errico · Feb 10, 2026
- In February Okta let AI agents reach its management APIs over MCP; Cloudflare added controls on MCP connections
Release tracker · · 7 vendors' release notes · Feb 2026
- The draft replacing RFC 8725 requires rejecting malformed JWTs but only recommends denial-of-service limits
Draft · · IETF datatracker · Yaron Sheffer et al. · Jan 12, 2026
- Cisco Duo's MCP beta lets agents register themselves; Duo calls client ID metadata documents the future
Article · · Cisco Duo · Colin Medfisch · Jan 7, 2026
- In January Okta's agent product was in early access; Auth0 and Microsoft made machine-to-machine features GA
Release tracker · · 7 vendors' release notes · Jan 2026
- A Shenzhen paper ties each AI agent to a verified owner and its exact code, at 15 to 40 seconds of proving a step
Paper · · arXiv · Zibin Lin et al. · Dec 19, 2025
- SIG Auth's storage-token fix needs driver changes
Video · · CNCF [Cloud Native Computing Foundation] · Uploaded November 24, 2025
- State Farm moved SPIRE from many token issuers to one, then hit an AWS key quota and a 15-character name limit
Video · · CNCF [Cloud Native Computing Foundation] · Nov 24, 2025
- A genuine TPM does not prove the server is LinkedIn's, so its SPIRE first checks the switch port and cabinet
Video · · CNCF [Cloud Native Computing Foundation] · Uploaded November 24, 2025
- For US banks to accept mobile driver's licenses, an OpenID Foundation group proposes five trust signals
Article · · OpenID Foundation · Serj Hallam · Nov 5, 2025
- Credential wallet draft makes holder binding a MUST; its own SD-JWT profile names no way to provide it
Draft · · IETF Datatracker · Leif Johansson et al. · Nov 4, 2025
- On its authors' tasks, an LLM check of agent scopes works at one tool and refuses most valid three-tool requests
Paper · · arXiv · Majed El Helou et al. · Oct 30, 2025
- 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
- An October SCIM draft gives agents their own records, with security still unfinished
Draft · · IETF Datatracker · Macy Abbey et al. · Oct 11, 2025
- A refresh-token thief who waits for the app to close slips past reuse detection
Article · · Auth0 Blog · Maria Vasilevskaya, Okta · Oct 8, 2025
- SPIFFE can authenticate an OAuth client; trust and registration still come first
Draft · · IETF Datatracker · Arndt Schwenkschuster et al. · Oct 3, 2025
- RFC 9864 deprecates ambiguous signature identifiers without prohibiting their use
Article · · IETF / RFC Editor · Michael B. Jones et al. · Oct 1, 2025
- Left to run a decentralized-ID check, AI agents finished as few as 4 runs in 100; one pair agreed to skip half
Paper · · arXiv · Sandro Rodriguez Garzon et al. · Oct 1, 2025
- One author's OpenID Connect for Agents proposal checks each delegation hop but makes the chain optional
Paper · · arXiv · Subramanya Nagabhushanaradhya · Sep 30, 2025
- OpenID's enterprise login draft adds tenant and session-expiry claims, and its security section is still blank
Article · · OpenID Foundation · Dick Hardt et al. · Sep 25, 2025
- CAEP 1.0 tells apps what changed after login and leaves almost every response to the receiver
Article · · OpenID Foundation · Tim Cappalli et al. · Aug 29, 2025
- Formal proofs find no new flaws in OpenID4VP over the Digital Credentials API, for a March 2025 draft
Article · · OpenID Foundation · Fabian Hauck et al. · Aug 20, 2025
- A 'provable' bound on agent-memory injection assumes independent layers and fills in none of its detection rates
Paper · · arXiv · Ken Huang et al. · Aug 17, 2025
- OpenID4VCI's third interop test: 87% of tested pairs passed, on counts the post does not reconcile
Article · · OpenID Foundation · Gail Hodges · Jul 24, 2025
- An OAuth client that never registers: the July 2025 draft makes its URL the identity and leaves most server checks as SHOULDs
Draft · · IETF Datatracker · Aaron Parecki et al. · Jul 23, 2025
- The July 2025 Cross App Access draft lets the IdP decide app-to-app access and treats agents as plain clients
Draft · · IETF Datatracker · Aaron Parecki et al. · Jul 2, 2025
- A Cloud Security Alliance paper wants AI agents on a new identity stack, with OAuth kept for simple cases
Paper · · arXiv (cs.CR preprint 2505.19301) / Cloud Security Alliance · Ken Huang et al. · May 25, 2025
- SAGA meters agent-to-agent contact through one registry; its speed and 300-million-agent figures are modelled
Paper · · arXiv · Georgios Syros et al. · Apr 27, 2025
- An A2A threat model flags the Agent Card as an attack path, and its example flow never passes on the user's identity
Paper · · arXiv · Idan Habler et al. · Apr 23, 2025
- An early MCP security checklist was never tested, and the protocol has since adopted only part of it
Paper · · arXiv · Vineeth Sai Narajala et al. · Apr 11, 2025
- A rogue OAuth server can collect a client's signed assertion and replay it at an honest server the client also uses
Article · · IACR Cryptology ePrint Archive · Pedram Hosseyni et al. · Apr 7, 2025
- Akeyless's CEO called just-in-time credentials the current focus and secretless an ideal that will take time
Podcast · · The Non-Human & AI Identity Podcast · The NHI Global Summit - Fireside Chat:Managing the full-lifecycle of Secrets & Non-Human Identities · First aired February 27, 2025 · Sep 4, 2025
- FAPI 2.0 issues only sender-constrained access tokens and leaves public clients out
Article · · OpenID Foundation · Daniel Fett et al. · Feb 22, 2025
- An MIT-led paper has the user sign the agent's delegation token and lets OpenID Connect vouch for both parties
Paper · · arXiv · Tobin South et al. · Jan 16, 2025
- RFC 9700 bans OAuth's password grant and loose redirect matching; binding access tokens stays a SHOULD
Article · · IETF (RFC Editor) · Torsten Lodderstedt et al. · Jan 2025
- In GNAP (RFC 9635), a client is known by its key, and tokens stay key-bound unless the server flags them bearer
Article · · RFC Editor · Justin Richer et al. · Oct 1, 2024
- DIF's Presentation Exchange lets a verifier ask for credentials, but not ask who is presenting them for whom
Article · · Decentralized Identity Foundation · Daniel Buchner et al. · Apr 25, 2024
- OpenID Connect Core says only issuer plus subject identify a user; email must never be the key
Article · · OpenID Foundation · Nat Sakimura et al. · Dec 15, 2023
- JWT sub and sub_id must identify the same subject
Article · · RFC Editor · Annabelle Backman et al. · Dec 6, 2023
- RFC 9470 lets an API demand a fresher or stronger sign-in, with no shared scale for 'stronger'
Article · · RFC Editor · Vittorio Bertocci et al. · Sep 8, 2023
- DPoP token binding depends on checks at the resource server
Article · · RFC Editor · Daniel Fett et al. · Sep 2023
- Rich Authorization Requests need API-specific permission rules
Article · · RFC Editor · Torsten Lodderstedt et al. · May 1, 2023
- RFC 9207 has OAuth servers name themselves in each response; clients must refuse a missing name only where they expect one
Article · · RFC Editor · Karsten Meyer zu Selhausen et al. · Mar 1, 2022
- RFC 9068 makes resource servers check a JWT access token's type and audience, and warns a client could pose as a user
Article · · RFC Editor · Vittorio Bertocci, Okta · Oct 1, 2021
- CIBA's human-approval step carries a short matching code, not a description of the action
Article · · OpenID Foundation · Gonzalo Fernandez Rodriguez et al. · Sep 1, 2021
- Signed OAuth requests can be bypassed until the server refuses unsigned ones, and RFC 9101 leaves that off by default
Article · · RFC Editor · Nat Sakimura et al. · Aug 1, 2021
- RFC 8725: the caller, not the JWT header, picks the algorithm, and "none" is discouraged, not banned
Article · · RFC Editor · Yaron Sheffer et al. · Feb 2020
- The token exchange RFC that agent designs cite says its record of earlier actors is not for access decisions
Article · · IETF Datatracker · Michael B. Jones et al. · Jan 1, 2020
- In draft FAPI, token binding did not make phished tokens useless: the attacker had the bound client spend them
Paper · · arXiv · Daniel Fett et al. · Jan 31, 2019
- RFC 8485 makes assurance values meaningful only within a named trust framework
Article · · RFC Editor · Justin Richer et al. · Oct 12, 2018
- Under RFC 8417 a security event token reports a fact, not an order, and the receiver decides what to do
Article · · RFC Editor · Phil Hunt et al. · Jul 2018
- RFC 8414 tells an OAuth client to discard server metadata whose issuer differs from the URL it started from
Article · · RFC Editor · Michael B. Jones et al. · Jun 1, 2018
- CWT leaves required claims to the application, including issuer, audience and expiry
Article · · RFC Editor / IETF · Michael B. Jones et al. · May 1, 2018
- RFC 8176 reports authentication methods separately from assurance classes
Article · · IETF / RFC Editor · Michael B. Jones et al. · Jun 1, 2017
- A 2017 security proof for OpenID Connect assumes fixes the spec did not yet require
Paper · · arXiv · Daniel Fett et al. · Apr 27, 2017
- RFC 7800 gives a JWT one place to name its holder's key, and leaves the proof to each protocol
Article · · RFC Editor · Michael B. Jones et al. · Apr 1, 2016
- Trying to prove OAuth 2.0 secure, a 2016 Trier paper found four attacks, the IdP mix-up among them
Paper · · arXiv · Daniel Fett et al. · Jan 6, 2016
- In 2015, OpenID Connect discovery let an attacker's server choose where a client sent the user's code
Paper · · arXiv · Christian Mainka et al. · Aug 18, 2015
- RFC 7591 asks OAuth servers to accept clients with no token, and calls their signed statements self-asserted
Article · · RFC Editor · Justin Richer et al. · Jul 9, 2015
- RFC 7519 makes "none" mandatory to implement and leaves refusing an unwanted algorithm as a SHOULD
Article · · RFC Editor · Michael B. Jones et al. · May 1, 2015
- Microsoft's 2023 keynote: workload standards should carry identity across service calls first, policy later
Video · · Identiverse - A CRA Resource · Recorded May 31, 2023 · Uploaded June 17, 2024
Related terms
What appears alongside OpenID Connect in the same briefings.
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.