IBAC
intent-based access control
Intent-based access control: a research design that checks each AI-agent tool call against an authorization store derived from the user's stated intent before the call runs.
In our briefings
The claim each of the newest briefings on IBAC would have you leave with.
Asked which identity and security controls are most likely to break first as agentic systems scale, John Yeoh, CSA's chief scientific officer, names lifecycle management. He cites CircleCI and Microsoft's Midnight Blizzard breach as poor lifecycle management of non-human identities, where keys for secrets were not rotated fast enough, and expects agents to fail faster because people now give them specs that spin up sub-agents. Access reviews that ignore intent and behaviour come next, then logging, in what he calls "a good domino of control collapse". Aizome's Amir Ofek answers the same question differently: applying controls built for non-human identities, which are deterministic rather than dynamic, is what will break first.
CSA's John Yeoh expects lifecycle management to be the first agent control to breakBI-FIND-2026-0113
Six voices say that knowing who an agent is and what it may access does not settle whether its next action is right, so the check has to look at what the agent is doing while it runs. Steven Rennick of Ciena says "we might know who the agent is and what it's allowed to do, but we really don't know how it's going to make those decisions," so trust has to become continuous and work at runtime. Sathvik Krishnamurthy, an investor with Security Leadership Capital speaking on Entrust's podcast, says identity and zero trust do not control an agent's behavior, and that a layer on top of them has to. Josh Woodruff of the Cloud Security Alliance and MassiveScale AI says zero trust checks who you are and what you can reach, and agents need one more check on every action: what are you doing right now, and is it allowed? Michael Davis of JPMorganChase says most detective controls only ask whether a tool call is allowed, and wants security to look at the actions themselves, not just the permissions. Sagar Batchu of Speakeasy retells an agent that deleted a production database with a valid full-access token: it "had valid credentials, but it didn't have valid intent." Dror Zelber of Radware wants each action compared with the agent's stated intent and goals, and blocked when they do not match. A seventh voice, Maia Iyer of IBM Research, describes the idea behind one such check, which could flag an outgoing call the user never asked for, and calls checks of this kind a soft layer for defense in depth. They differ on how much intent can carry. Rennick and Zelber would judge each action against the agent's intent. Woodruff says the checkpoint must be able to refuse an action whatever the intent was, because an agent with the right intent can still do damage. Rennick adds that judging actions by their outcomes as well is the desired end state, but he does not think the technology is there yet.
Knowing who an agent is and what it may reach is not enough, six voices say: check each action as it runsBI-FIND-2026-2330
A person's approved request is not, by itself, authority an agent can be held to, McGuinness argues, and relatively few deployments perform the step that would make it one. That step converts approved intent into a bounded authority artifact an enforcement point can evaluate. By his account, many systems instead fall into one of five patterns: scope enumeration, app-local mapping, LLM-generated requests, broad standing authority with prompt discipline, or a formal policy language. He says the most common land in the third or fourth, and that where nothing makes the conversion, the agent's own judgment becomes the effective policy engine. This is one author's argument, but two production designs we have read fit it. Uber's agent tokens carry the full chain of who asked through every hop, and Sierra's gateway runs interactive agent work as the user, with that person's permissions. In our reading, neither records what the person approved.
A user's approved request is not yet authority an agent can be held to, and most deployments skip that stepBI-FIND-2026-0202
›Cited in 9 briefings
Every briefing that discusses this term, newest first.
- CSA's John Yeoh expects lifecycle management to be the first agent control to break
Podcast · · The Non-Human & AI Identity Podcast · NHI & AI Identity Summit : The Next 24 Months · Sep 7, 2026
- Knowing who an agent is and what it may reach is not enough, six voices say: check each action as it runs
Field signals · · 14 voices from the field · Dec 2025 to Nov 2026
- A user's approved request is not yet authority an agent can be held to, and most deployments skip that step
Article · · Control Plane by Karl McGuinness · Karl McGuinness · Mar 17, 2026
- OAuth token exchange does not guarantee a sub-agent stays inside its parent's access
Article · · Control Plane by Karl McGuinness · Karl McGuinness · Mar 18, 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
- Registering an MCP server on Bifrost ran its command, and the admin API that allowed it shipped with no password
Article · · WorkOS Blog · Maria Paktiti · Sep 29, 2026
- The IBAC paper's strict mode blocked every prompt injection, and its agent finished only a third of its tasks
Article · · ibac.dev · Jordan Potti · Mar 3, 2026
- Huawei researchers add agent delegation to an existing permission model and prove each action traces to a human
Paper · · arXiv · Amjad Ibrahim et al. · Jun 2, 2026
- Checking each agent call against the user's own task stopped all 634 off-task calls the authors wrote
Paper · · arXiv · Reshabh K Sharma et al. · Mar 17, 2026
Related terms
What appears alongside IBAC in the same briefings.