[论文解读] Deployment Corrections: An incident response framework for frontier AI models

J.N. O'Brien, Shaun Ee+1 more|arXiv (Cornell University)|Sep 30, 2023

Scientific Computing and Data Management被引用 5

一句话总结

本文提出了一套针对前沿人工智能模型的结构化事件响应框架，使开发者能够在模型部署后通过受控访问和预定义响应动作纠正危险行为。该框架借鉴网络安全实践，提出了一套部署修正工具包和治理框架，以主动管理高风险人工智能系统可能引发的灾难性风险。

ABSTRACT

A comprehensive approach to addressing catastrophic risks from AI models should cover the full model lifecycle. This paper explores contingency plans for cases where pre-deployment risk management falls short: where either very dangerous models are deployed, or deployed models become very dangerous. Informed by incident response practices from industries including cybersecurity, we describe a toolkit of deployment corrections that AI developers can use to respond to dangerous capabilities, behaviors, or use cases of AI models that develop or are detected after deployment. We also provide a framework for AI developers to prepare and implement this toolkit. We conclude by recommending that frontier AI developers should (1) maintain control over model access, (2) establish or grow dedicated teams to design and maintain processes for deployment corrections, including incident response plans, and (3) establish these deployment corrections as allowable actions with downstream users. We also recommend frontier AI developers, standard-setting organizations, and regulators should collaborate to define a standardized industry-wide approach to the use of deployment corrections in incident response. Caveat: This work applies to frontier AI models that are made available through interfaces (e.g., API) that provide the AI developer or another upstream party means of maintaining control over access (e.g., GPT-4 or Claude). It does not apply to management of catastrophic risk from open-source models (e.g., BLOOM or Llama-2), for which the restrictions we discuss are largely unenforceable.

研究动机与目标

• 填补前沿人工智能模型在部署后变得具有危险能力但缺乏风险管理的空白。
• 为人工智能开发者开发一种实用且可操作的响应机制，以纠正模型部署后的危险行为。
• 确保部署修正措施正式整合进人工智能开发与治理流程。
• 推动人工智能行业内事件响应实践的标准化，以提升安全性和问责性。
• 通过保持开发者对访问权限和响应动作的控制，支持前沿人工智能模型的负责任部署。

提出的方法

• 将网络安全领域的事件响应方法论适配至前沿人工智能部署的语境。
• 定义一组部署修正措施——如访问权限撤销、模型回滚和行为限制——作为核心响应动作。
• 建立人工智能开发者可预先准备、测试并实施部署修正的治理框架。
• 强调人工智能开发组织内设立专门事件响应团队及正式化流程的必要性。
• 建议将部署修正视为下游用户和监管机构认可且允许的合法行动。
• 倡导行业协作，推动在前沿人工智能系统间标准化部署修正协议。

实验结果

研究问题

• RQ1 人工智能开发者如何在模型部署后有效应对前沿人工智能模型的危险行为？
• RQ2 一旦模型上线并表现出有害能力，可以采取哪些具体措施来减轻灾难性风险？
• RQ3 其他高风险行业的事件响应实践如何适配前沿人工智能所面临的独特挑战？
• RQ4 为确保部署修正的及时与有效，需要哪些组织与治理结构？
• RQ5 如何在人工智能行业内实现部署修正措施的标准化，以确保一致性和问责性？

主要发现

• 针对前沿人工智能模型的全面事件响应框架可显著降低模型部署后的灾难性伤害风险。
• 部署修正措施——如访问控制、模型回滚和行为限制——是可行且必要的后部署风险管控工具。
• 保持开发者对模型访问的控制权，是实现有效部署修正的关键。
• 设立专门的事件响应团队并建立正式流程，对及时可靠地执行部署修正至关重要。
• 在人工智能行业内实现部署修正实践的标准化，是确保一致性、问责性和可扩展性的必要条件。
• 该框架不适用于开源模型，因为在这些模型中访问控制无法强制执行，凸显了其在适用范围上的关键局限。