🤖 AI Summary
Existing safety mechanisms struggle to constrain the option-generation and presentation authority of autonomous agents in high-stakes, regulated environments, often leading to irreversible decision risks. This work proposes a novel governance architecture that shifts the focus from intent alignment to the control of “choice power.” By structurally decoupling cognition, selection, and action, and integrating mechanisms such as external candidate generation (CEFL), controlled reducers, commit-reveal entropy isolation, rationale validation, and explicit fault circuit breakers, the framework enforces quantifiable bounds on an agent’s causal power. Empirical evaluations in financial adversarial settings demonstrate that the approach effectively prevents deterministic outcome manipulation, significantly reduces choice concentration, enhances narrative diversity and governance transparency, and preserves the agent’s original reasoning capabilities.
📝 Abstract
Autonomous agentic systems are increasingly deployed in regulated, high-stakes domains where decisions may be irreversible and institutionally constrained. Existing safety approaches emphasize alignment, interpretability, or action-level filtering. We argue that these mechanisms are necessary but insufficient because they do not directly govern selection power: the authority to determine which options are generated, surfaced, and framed for decision. We propose a governance architecture that separates cognition, selection, and action into distinct domains and models autonomy as a vector of sovereignty. Cognitive autonomy remains unconstrained, while selection and action autonomy are bounded through mechanically enforced primitives operating outside the agent's optimization space. The architecture integrates external candidate generation (CEFL), a governed reducer, commit-reveal entropy isolation, rationale validation, and fail-loud circuit breakers. We evaluate the system across multiple regulated financial scenarios under adversarial stress targeting variance manipulation, threshold gaming, framing skew, ordering effects, and entropy probing. Metrics quantify selection concentration, narrative diversity, governance activation cost, and failure visibility. Results show that mechanical selection governance is implementable, auditable, and prevents deterministic outcome capture while preserving reasoning capacity. Although probabilistic concentration remains, the architecture measurably bounds selection authority relative to conventional scalar pipelines. This work reframes governance as bounded causal power rather than internal intent alignment, offering a foundation for deploying autonomous agents where silent failure is unacceptable.