OpenAI与Anthropic如何平衡金融智能体的自主与安全
原标题：Crossing the Threshold: How OpenAI and Anthropic Approach Autonomy and Safety in Financial Agents
AI 摘要
在 AI Engineer New York 大会上，OpenAI、Anthropic、Two Sigma 和 Bloomberg 的演讲者讨论了金融智能体的自主性与安全平衡。OpenAI 的 Joseph Kim 提出“有用性阈值”概念，预测 2027 年智能体可完成投资想法生成和交易；Two Sigma 的 John Feminella 介绍了可组合控制措施和自动模式分类器以减少审批疲劳；Anthropic 的 Belinda Neal 和 Keon Shahab 强调为智能体建立身份、权限、沙箱和审批信任层；Bloomberg 的 Andrey Rybka 则主张在模型外部实施确定性控制，如范围化身份、工具调用控制、审计追踪和终止开关。
正文节选
One of the key themes at the AI Engineer New York conference this week was how to balance agent autonomy with security and safety. Both frontier labs and financial firms focused on this issue in their presentations this week. The main lesson: safe autonomy depends on where you put controls and human judgment, rather than how many restrictions you impose. Sponsor: Build shopping into your agent with shop.dev Adding commerce to an agent usually means rebuilding search and checkout from scratch. sh