Papers/2608.16891
🧪 Test?View on arXiv

Runtime Governance for Agentic AI: Action-Boundary Control with Trusted Provenance and Fail-Closed Execution

Not provided in the abstract

runtime governanceagentic AIsafety mechanismspolicy evaluation
2608.16891
Builder Relevance
80%
2h ago

Abstract

The paper introduces Aegis, a runtime governance system that prevents risky operational side effects from agentic AI systems by mediating tool action proposals through a trusted decision layer.

Reality Card

Core Claim

Aegis effectively prevented all observed risky proposals from becoming governed side effects in a controlled sandbox environment.

Method / Result

In 2,100 Aegis-governed rows, the system recorded zero governed mock-tool applications and zero governed risky side-effect completions.

Limitations

The results do not prove general autonomous-agent safety and are limited to the evaluated sandbox corpus.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.
← Back to all papers