Structural Enforcement of Goal Integrity in AI Agents via Separation-of-Powers Architecture

TL;DR

This paper introduces the PEA architecture, a system-level design that enforces goal integrity in AI agents through separation of powers, cryptographic tokens, and formal verification, addressing agentic misalignment risks.

Contribution

The paper presents a novel system architecture with five core components that structurally enforce safety and goal integrity in autonomous AI agents.

Findings

PEA architecture effectively enforces goal integrity even under adversarial conditions.

Cryptographically anchored intent tracking ensures traceability of agent actions.

Formal verification proves goal safety under potential system compromises.

Abstract

Recent evidence suggests that frontier AI systems can exhibit agentic misalignment, generating and executing harmful actions derived from internally constructed goals, even without explicit user requests. Existing mitigation methods, such as Reinforcement Learning from Human Feedback (RLHF) and constitutional prompting, operate primarily at the model level and provide only probabilistic safety guarantees. We propose the Policy-Execution-Authorization (PEA) architecture, a "separation-of-powers" design that enforces safety at the system level. PEA decouples intent generation, authorization, and execution into independent, isolated layers connected via cryptographically constrained capability tokens. We present five core contributions: (C1) an Intent Verification Layer (IVL) for ensuring capability-intent consistency; (C2) Intent Lineage Tracking (ILT), which binds all executable intents to the originating user request via cryptographic anchors; (C3) Goal Drift Detection, which rejects semantically divergent intents below a configurable threshold; (C4) an Output Semantic Gate (OSG) that detects implicit coercion using a structured $K \times I \times P$ threat calculus (Knowledge, Influence, Policy); and (C5) a formal verification framework proving that goal integrity is maintained even under adversarial model compromise. By shifting agent alignment from a behavioral property to a structurally enforced system constraint, PEA provides a robust foundation for the governance of autonomous agents.