Governing Dynamic Capabilities: Cryptographic Binding and Reproducibility Verification for AI Agent Tool Use

TL;DR

This paper introduces a cryptographic framework for verifying the capabilities, behavior, and interactions of AI agents to ensure security, provenance, and reproducibility in multi-agent systems.

Contribution

It defines new governance requirements, proves structural theorems for verification, and validates cryptographic instantiations with practical performance and security guarantees.

Findings

Verification overhead is less than 0.02%.

Reproducibility variance is 5.8x across models.

All attack scenarios are detected with zero false positives.

Abstract

AI agents dynamically acquire tools, orchestrate sub-agents, and transact across organizational boundaries, yet no existing security layer verifies what an agent can do, whether it executed what it claims, or what happened in a multi-agent interaction. We trace this gap to the capability-context separation: inside a transformer, tool definitions and user context are indistinguishable tokens, but at the orchestration layer they have fundamentally different security semantics. Existing frameworks conflate the two, enabling silent capability escalation and leaving interactions without verifiable provenance. From this principle we derive three Agent Governance Requirements: capability integrity (G1), behavioral verifiability (G2), and interaction auditability (G3), defining what a governed agent ecosystem must enforce, independent of how. We prove two structural results: the Chain Verifiability Theorem (one unverifiable interior agent breaks end-to-end verification for all downstream nodes) and the Bounded Divergence Theorem (replay-based verification yields a probabilistic safety certificate, epsilon <= 1 - alpha^{1/n}). We validate with two crypto-agnostic instantiations -- basic (Ed25519, SHA-256; 97 us verify) and enhanced (BBS+ selective disclosure, Groth16 DV-SNARK; 13.8 ms) -- both satisfying nine security properties. A reproducibility study (9 models, 7 providers) reveals 5.8x variance in inference determinism, connecting model characteristics to governance architecture. End-to-end evaluation over 5-20 agent pipelines confirms <0.02% overhead and detection of all attack scenarios with zero false positives.