Anthropic、Google、Microsoft 和 OpenAI 同意让网站退出 AI 训练
Anthropic, Google, Microsoft, and OpenAI Agree to Let Sites Opt Out of AI Training
Anthropic、Google、Microsoft 和 OpenAI 同意支持并实施由 IETF 工作组制定的退出机制，让网站所有者可以就内容用于生成式 AI 训练表达偏好，并与搜索索引和归档区分开。OpenAI 称这是一项自愿机制，可让发布者一次性向多家参与厂商传达同一偏好；Microsoft 表示发布者不必在 Bing 可发现性和模型使用其作品之间二选一。Anthropic 已实施该机制，Google、Microsoft 和 OpenAI 将在 2027 年实施，该机制完成后将作为国际标准提案。
推荐理由
AI 评分 70
IETF 工作组推出的统一退出信号让网站可一次性向多家厂商表达训练偏好，Anthropic 已落地，其余三家排期 2027 年。
主题
正文
Some of the world’s biggest AI companies have agreed to let website owners opt out of AI model training. Unfortunately, this move comes years after all of them stole content at a scale never before seen, making the change largely symbolic.
Anthropic, Google, Microsoft, and OpenAI have decided to support and implement the new system.
“Publishers should not have to choose between being discoverable in Bing and deciding how generative models use their work,” Microsoft principal product manager Krishna Madhavan writes. “Today, we are offering newer, simpler, and clearer preferences that allow you to make choices about training, grounding, or both while remaining eligible for Bing search results.”
“OpenAI will honor a common signal … that lets websites express their preferences about the use of their content for foundational generative AI training,” the OpenAI announcement explains. “This voluntary effort makes it easier for publishers and website operators to communicate the same preference to multiple participating providers at once.”
The opt out system was created by a working group at the Internet Engineering Task Force (IETF) and it will be proposed as an international standard once it’s complete. It allows a website to communicate to all AI content scrapers that they wish to opt out of this behavior separately from search indexing and archiving. Anthropic has implemented it already, but Google, Microsoft, and OpenAI will do so in 2027.