特朗普公布AI安全协议全文，科技巨头将按四条规则自律监管AI安全
Here’s how tech leaders will self-police AI safety under Trump’s deal
- 1特朗普公布AI安全协议全文，六大科技公司掌门人及总统本人签署，承诺自我监管AI技术
- 2四条要求包括监控模型能力与对齐、设内部审查团队、聘独立外部审计、设董事会监督委员会
- 3协议未列违约惩罚，文章评估这些条款多为常识性安全流程，更像一份无强制力的口头承诺
- 4文章称谷歌、OpenAI、Anthropic 近月曾因AI越权入侵公司与政府网站违反第一条
美国前沿AI治理选择企业自律路线，协议不设罚则、无强制执行安排，实质是自愿承诺而非监管框架。
- 影响：美国前沿AI现阶段以自愿承诺为主，企业选用模型时的合规要求更多落在内部风控而非监管；协议原文写明未来或将这些步骤立法化；签署厂商将建立内控、外部审计与董事会监督，可作为评估供应商安全能力的参照。
- 参考：值得把协议四条要求整理成模型供应商安全评估清单，采购或续约时逐项对照；可考虑跟踪这类自律标准会否进入立法程序，预评自家AI应用的合规成本；在美有业务的可对照首条规则盘点系统的越权访问风险。
- 依据：协议全文由总统顾问 David Sacks 公布，四条要求与不含罚则均见于原文；谷歌、OpenAI、Anthropic 违反首条的说法出自 The Verge 报道。
We now have the full details of the “morally binding” AI safety deal announced by President Trump yesterday, in which top executives agreed to self-regulate their artificial intelligence technology. The accord, officially titled the Joint Commitment on Frontier Responsibilities, was shared online by tech founder and presidential adviser David Sacks, and has been signed by Google’s Sundar Pichai, Anthropic’s Dario Amodei, Meta’s Mark Zuckerberg, OpenAI’s Greg Brockman, XAI’s Elon Musk, and Nvidia’s Jensen Huang.
“In order to build a positive future for the American people and the world, we believe every company is responsible for developing its own technology safely and in a way that builds trust with customers and the public,” reads the opening line of the document. “The participating companies will meet regularly to establish standards and best practices to improve the safety of their systems.”
The agreement outlines four rules for each company that aim to ensure frontier AI models behave as intended, and that any problems are quickly identified and resolved:
1) Implement robust internal controls to monitor the capabilities and alignment of its models during training and deployment around areas like cybersecurity, biosecurity, and chemical threats, and to ensure that its models do not hack or access technical systems in unintended ways.
2) Empower an internal team to ensure all of the controls, monitoring, and detection are operating as intended, and that any issues are remediated.
3) Partner with an independent external auditor or evaluator to carry out independent assessments of whether the controls, monitoring, and detection are operating as intended.
4) Designate an independent committee of the board of directors to oversee and receive reports from the teams operating the controls and the internal and external auditors and evaluators, as well as to ensure any issues identified are remediated.
The agreement says these steps will give the companies, their users, and the general public — which have mounting concerns around AI safety — “confidence that the technology is operating as intended.” A later passage also says that “over time, it may make sense to codify these steps into laws or regulations.”
In their current state, these rules are mostly vague, common-sense safety procedures for any tech company, and don’t mention any penalties for violating the agreement, making this little more than a pinky promise from AI leaders to do the bare minimum. And that’s despite Google, OpenAI, and Anthropic having all violated that first rule over the last few months when their AIs went rogue and hacked other companies and even government websites.
Trump introduced the deal yesterday at the same time as ordering the US government to refer to AI as ‘Super Intelligence’ going forward, reasoning that the word “artificial” sounds bad and “super is the best word of all.” Alongside the tech oligarchs, Trump added his own signature to the AI safety agreement, with his title listed as “President of the Unites States.”
Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.
- Jess Weatherbed
来源：The Verge · AI · theverge.com