Skip to content

Topic

Agent Identity and Authorization

How AI agents are authenticated and authorized: credential and token scoping, agent identity providers, and authority propagation across agent actions.

Current stories

securityOne report1 publisher

ServiceNow issues each AI agent a signed credential that defines what it may do at runtime

ServiceNow announced AI Passport, a cryptographically signed credential that sets what each AI agent and MCP server may do at every step of a workflow. The prevention claims are ServiceNow's own description of a design.

Reality

Evidence25
Adoption
Not scored
Hype gap+40
Incentives70
Confidence60
buildOne report1 publisher

kagent 0.10 ships its RFC 8707 token scoping as an empty string you fill in yourself

The GA release registers KAGENT_STS_RESOURCE and KAGENT_STS_AUDIENCE with descriptions that name the RFC, then defaults both to empty, and whether that default is an exposure depends on what your own STS hands back.

Publishers:dev.to

Reality

Evidence68
Adoption22
Hype gap+12
Incentives35
Confidence60
productOne report1 publisher

CrowdStrike wants to be the identity provider your AI agents register with

The new Agentic IdP registers every agent Falcon Guardian finds, issues it a cryptographic identity and brokers short-lived tokens instead of credentials, which quietly makes your discovery coverage the thing that decides the rollout.

Publishers:siliconangle.com

Reality

Evidence36
Adoption14
Hype gap+34
Incentives84
Confidence57
investOne report1 publisher

Lianlian puts agent-wallet controls in a stock filing, on a 24-basis-point take rate

Lianlian DigiTech's Hong Kong filing names agent identity, spending limits and traceability as the product. Its cross-border economics say that control layer has to cost almost nothing.

Reality

Evidence55
Adoption30
Hype gap+32
Incentives68
Confidence52
buildOne report1 publisher

AWS's one-minute test for agent access is really a test of where the answer lives

Amazon says you should be able to name every AI agent touching customer data in under a minute. Its own worked example explains why most teams cannot: the record is a file on a laptop.

Publishers:aws.amazon.com

Reality

Evidence34
Adoption
Not scored
Hype gap+28
Incentives82
Confidence46
productOne report1 publisher

Agents Are Not Microservices With an LLM Attached, and the Retrofit Never Arrives

A devops.com essay argues durability, per-step identity and decision-level observability belong in the agent runtime, not layered on later. The failure economics support the claim.

Publishers:devops.com

Reality

Evidence24
Adoption
Not scored
Hype gap+34
Incentives55
Confidence38
securityOne report1 publisher

OWASP now names seventeen agentic threats, and the control unit is the whole run

Version 1.1 of the Agentic Security Initiative's guide enumerates T1 through T17, up from fifteen. Cite the version, and stop reviewing agents one tool call at a time.

Publishers:orca.security

Reality

Evidence52
Adoption
Not scored
Hype gap+12
Incentives74
Confidence48