How to Catch a GPU: A Taxonomy of Verification and Enforcement Mechanisms for International AI Agreements

• URL: http://arxiv.org/abs/2607.22619v1
• Date: Tue, 16 Jun 2026 10:51:52 GMT
• Title: How to Catch a GPU: A Taxonomy of Verification and Enforcement Mechanisms for International AI Agreements
• Abstract summary: We propose a taxonomy based on the principle that wherever sufficient capacity exists to violate an agreement, it must be under a control regime.<n>This decomposes the problem of ensuring compliance with the agreement into preventing uncontrolled resource acquisition, detecting all capacity outside the control regime, and preventing escape from the control regime.<n>Applying this taxonomy to existing proposals reveals that more focus is placed on preventing escape from the control regime, while preventing resource acquisition and detecting all capacity outside the control regime receive less attention.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Several international agreements have been proposed to regulate frontier AI development in response to catastrophic risks. However, there is no structured way to evaluate whether these proposals are enforceable, to assess where they might fail in practice, or to determine which combination of policies is most effective. We propose a taxonomy based on the principle that wherever sufficient capacity exists to violate an agreement, it must be under a control regime. This decomposes the problem of ensuring compliance with the agreement into preventing uncontrolled resource acquisition, detecting all capacity outside the control regime, and preventing escape from the control regime. Existing proposals consist of individual policies that address one or more of these sub-problems. Because the compute required for dangerous capabilities may decrease over time, more actors can violate an agreement and enforcement of these policies becomes harder. We define the enforcement breaking point as the FLOP-threshold or equivalent metric at which a policy loses its effectiveness in solving the sub-problem, and introduce a set of factors to assess how and why this breakdown occurs. This reveals which of the three sub-problems any given proposal adequately addresses, and where enforceability breaks down first. Applying this taxonomy to existing proposals reveals that more focus is placed on preventing escape from the control regime, while preventing resource acquisition and detecting all capacity outside the control regime receive less attention. By making these gaps explicit, this taxonomy can help researchers and policymakers prioritize future enforcement and verification efforts.

Related papers

• Right-to-Act: A Pre-Execution Non-Compensatory Decision Protocol for AI Systems [0.0]
  We introduce the Right-to-Act protocol, a deterministic, non-compensatory pre-execution decision layer that evaluates whether an AI-generated decision is permitted to be realized.<n>Unlike compensatory systems, where high-confidence signals can override failed conditions, the proposed framework enforces strict structural constraints.
  arXiv  Detail & Related papers  (2026-04-27T08:09:12Z)
• Verification of Robust Properties for Access Control Policies [51.736723807086385]
  Existing methods for verifying access control policies require the policy to be complete and fully determined before verification can proceed.<n>We introduce robust property verification: the problem of determining what a policy's structure commits it to regardless of how pending decisions are resolved and regardless of subsequent extension.<n>We show that despite quantifying universally over all possible policy extensions the judgment reduces to proof search in a second-order logic programming language.
  arXiv  Detail & Related papers  (2026-03-13T17:14:38Z)
• Autonomous Agents and Policy Compliance: A Framework for Reasoning About Penalties [1.2891210250935148]
  This paper presents a logic programming-based framework for policy-aware autonomous agents that can reason about potential penalties for non-compliance.<n>Our framework extends Gelfond and Lobo's Authorization and Obligation Policy Language (AOPL) to incorporate penalties.<n>Our method ensures well-formed policies, accounts for policy priorities, and enhances explainability by explicitly identifying rule violations.
  arXiv  Detail & Related papers  (2025-12-03T16:29:09Z)
• The Loss of Control Playbook: Degrees, Dynamics, and Preparedness [39.39076397908963]
  This report addresses the absence of an actionable definition for Loss of Control (LoC) in AI systems by developing a novel taxonomy and preparedness framework.<n>We propose a graded LoC taxonomy, based on the metrics of severity and persistence, that distinguishes between Deviation, Bounded LoC, and Strict LoC.<n>We put forward a plan to maintain preparedness and prevent the occurrence of LoC outcomes should a state of societal vulnerability be reached.
  arXiv  Detail & Related papers  (2025-11-19T20:10:39Z)
• Limits of Safe AI Deployment: Differentiating Oversight and Control [0.0]
  "Human oversight" risk codifying vague or inconsistent interpretations of key concepts like oversight and control.<n>This paper undertakes a targeted critical review of literature on supervision outside of AI.<n>Control aims to prevent failures, while oversight focuses on detection, remediation, or incentives for future prevention.
  arXiv  Detail & Related papers  (2025-07-04T12:22:35Z)
• Learning Deterministic Policies with Policy Gradients in Constrained Markov Decision Processes [59.27926064817273]
  We introduce an exploration-agnostic algorithm, called C-PG, which enjoys global last-iterate convergence guarantees under domination assumptions.<n>We empirically validate both the action-based (C-PGAE) and parameter-based (C-PGPE) variants of C-PG on constrained control tasks.
  arXiv  Detail & Related papers  (2025-06-06T10:29:05Z)
• Rethinking State Disentanglement in Causal Reinforcement Learning [78.12976579620165]
  Causality provides rigorous theoretical support for ensuring that the underlying states can be uniquely recovered through identifiability. We revisit this research line and find that incorporating RL-specific context can reduce unnecessary assumptions in previous identifiability analyses for latent states. We propose a novel approach for general partially observable Markov Decision Processes (POMDPs) by replacing the complicated structural constraints in previous methods with two simple constraints for transition and reward preservation.
  arXiv  Detail & Related papers  (2024-08-24T06:49:13Z)
• Last-Iterate Global Convergence of Policy Gradients for Constrained Reinforcement Learning [62.81324245896717]
  We introduce an exploration-agnostic algorithm, called C-PG, which exhibits global last-ite convergence guarantees under (weak) gradient domination assumptions. We numerically validate our algorithms on constrained control problems, and compare them with state-of-the-art baselines.
  arXiv  Detail & Related papers  (2024-07-15T14:54:57Z)
• Do No Harm: A Counterfactual Approach to Safe Reinforcement Learning [5.862025534776996]
  Reinforcement Learning for control has become increasingly popular due to its ability to learn rich feedback policies that take into account uncertainty and complex representations of the environment. In such methods, if agents are in, or must visit, states where constraint violation might be inevitable, it is unclear how much they should be penalized. We address this challenge by formulating a constraint on the counterfactual harm of the learned policy compared to a default, safe policy. In a philosophical sense this formulation only penalizes the learner for constraint violations that it caused; in a practical sense it maintains feasibility of the optimal control problem.
  arXiv  Detail & Related papers  (2024-05-19T20:33:21Z)
• Verification of indefinite-horizon POMDPs [63.6726420864286]
  This paper considers the verification problem for partially observable MDPs. We present an abstraction-refinement framework extending previous instantiations of the Lovejoy-approach.
  arXiv  Detail & Related papers  (2020-06-30T21:01:52Z)
• Hierarchical Adaptive Contextual Bandits for Resource Constraint based Recommendation [49.69139684065241]
  Contextual multi-armed bandit (MAB) achieves cutting-edge performance on a variety of problems. In this paper, we propose a hierarchical adaptive contextual bandit method (HATCH) to conduct the policy learning of contextual bandits with a budget constraint.
  arXiv  Detail & Related papers  (2020-04-02T17:04:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.