OpenAI、谷歌、微软和Anthropic等主要人工智能公司宣布，将采用一项由互联网工程任务组（IETF）提议的新标准，旨在简化出版商拒绝AI训练数据抓取的流程 1。该标准的核心机制是在网站现有的robots.txt文件中引入共享的“政策令牌”（policy token）信号 1。
通过这一新方案，出版商只需在robots.txt中设置统一的偏好，即可跨多家AI公司应用其对AI训练数据的抓取限制，而无需针对每家公司单独进行配置 1。该提案由Anthropic的技术标准负责人Kevin Kelley提出 1。
Leading artificial intelligence companies, including OpenAI, Google, Microsoft, and Anthropic, have announced their support for a new industry standard designed to streamline the process of opting out of AI data scraping. 1 The proposal was introduced by Kevin Kelley, Head of Technical Standards at Anthropic, and is being advanced under the framework of the Internet Engineering Task Force (IETF). 1 This initiative aims to unify how publishers manage their preferences regarding AI training data across multiple tech giants. 1
The core mechanism involves introducing shared "policy tokens" within existing robots.txt files on websites. 1 By implementing these tokens, publishers can set their AI training preferences once, which will then be recognized by all participating AI companies, eliminating the need for separate, company-specific opt-out configurations. 1
评论
还没有评论，欢迎留下第一条。