OpenAI, Google, Microsoft and Anthropic jointly promote AI data scraping opt-out standard
English text is being prepared.
OpenAI, Google, Microsoft and Anthropic announced that they will adopt a new standard that allows content publishers to declare their content usage preferences to multiple AI companies at once via robots.txt, without having to set them individually for each company. The standard was proposed to the IETF by Anthropic technical lead Kevin Kelley. Anthropic has already adopted it first, OpenAI plans to implement it within the next few months, and Google and Microsoft are expected to support it next year. Meta spokesperson Andy Stone said he supports establishing such a universal signal, but did not specify an implementation plan.
Why read this
AI score 68
The new standard replaces per-company declarations with a unified statement in robots.txt; readers can learn about each company's implementation timeline and the differences in Meta's stance.
Topics
Body
OpenAI、谷歌、微软和 Anthropic 联手推行 AI 数据抓取退出标准
2026-11-01 22:19 IT之家 - 浩渺
感谢IT之家网友 咩咩洋 的线索投递！
IT之家 11 月 1 日消息，OpenAI、谷歌、微软以及 Anthropic 近日宣布将采用一项新标准，内容发布方可以借助该标准选择拒绝 AI 抓取自家网站内容。借助这项管控机制，内容发布方可以一次性向多家 AI 企业统一申明自身针对 AI 模型训练的内容使用偏好，而不必为每家企业单独设置规则。
该标准由 Anthropic 技术负责人 Kevin Kelley 向互联网工程任务组（IETF）提出，并将集成在网站的 robots.txt 文件中。
长期以来，robots.txt 文件用于告知网络爬虫哪些网站内容允许抓取、哪些禁止抓取，适用场景包含搜索索引、网页归档以及 AI 模型训练。目前，网站必须向每一家 AI 企业分别单独声明自己的 AI 训练相关内容使用偏好。在此背景下，各类 AI 爬虫持续抓取网站内容、将网页素材输入 AI 模型训练，内容发布方正为此焦头烂额。
IT之家注意到，目前，网站需向每家公司单独声明训练偏好，新标准旨在简化这一流程。Anthropic 已率先采纳，OpenAI 计划在未来几个月内落实，谷歌和微软则预计于明年支持。Meta 发言人 Andy Stone 也表示支持建立此类通用信号，但未明确具体落地计划。