Threats are stopped at the boundary closest to the consequence

BeaconGuard assumes the AI or agent can be wrong, compromised, manipulated, or outside the authority required for a consequential action. AI reasoning, identity alone, or broad tool permission is therefore never treated as the final security boundary.

BeaconGuard threat model showing identity and delegation confusion, access mistaken for authority, out-of-policy AI actions, action substitution, grant replay or reuse, release tampering, evidence alteration, and downstream bypass around the exact-action authorization boundary.
Threats are addressed at the boundary closest to the consequence: identity and delegation, invocation, exact-action authorization, execution authority, policy authority, evidence, and deployment placement.

Primary threat matrix

ThreatBeaconGuard design response
Identity or delegation confusionValidate required provenance, bounds, tenant/environment, expiry/revocation, and fail closed when trusted context cannot be established.
Access permission treated as blanket action authorityPreserve the separation between invocation access and a later exact consequential-action authorization decision.
AI proposes an out-of-policy side effectTreat the output as an untrusted proposal and evaluate the exact action under customer-approved governing policy.
Action substitution after ALLOWBind authorization to a deterministic exact-action identity and recompute that identity at the executor.
Execution-grant replay or reuseConstrain grant validity and consume one-time execution authority once; reuse attempts fail closed.
Release tampering or unauthorized activationRequire signed release artifacts, verification, and explicit activation before policy may govern production.
Evidence alteration or suppressionPreserve integrity-protected authorization evidence and verification context; durability and retention remain provider-dependent.
Downstream bypassKeep BeaconGuard on the execution path and require the executor to reject missing or invalid action authority.

Protected assets

Security assumptions and residual risk

BeaconGuard does not claim absolute security. Important residual dependencies remain:

Related technical references