- Published on
Meta revises business terms, sparking 'AI training' controversy, now deletes relevant wording
- Authors
- Name
- DeepHorus
IT Home, October 18 news: Meta recently updated its business terms of service, stating that it has the right to “use any content from [business account] domains, web pages, or URLs to provide, develop, improve, research, train, and maintain” the integrity of Meta products and services. The clause quickly attracted outside attention. At the heart of the controversy is whether Meta can use it to scrape the websites of content publishers that have Meta business accounts and use the relevant content to train artificial intelligence models.
Legal experts interviewed by Nieman Lab pointed out that this is a “striking clause.” They said the clause could give Meta the authority to scrape any content publisher website that has a Meta business account and use the relevant content to train its artificial intelligence models. This interpretation sparked concerns about the flow of data from content publishers’ websites. Expressions such as business account, domain, and URL became key points in the controversy.
However, a Meta spokesperson insisted that the updated terms do not mean Meta’s way of crawling the web has changed. This response attempted to confine the controversy to the space between the wording of the terms and actual crawler behavior. Nevertheless, the relevant wording still attracted attention from media and industry observers. According to The Verge, the clause has now been officially deleted.
Meta said in a statement: “We have listened to feedback from all sides and would like to clarify: this update is a ‘consolidation’ of existing advertising terms, not an expansion of crawling behavior.” The statement also said that the revised terms now explicitly limit crawling and data collection to “advertising and business products.” Meta also reiterated that webmasters can still control crawler permissions through robots.txt.
Judging from Meta’s clarification, this update was described as a consolidation of existing advertising terms, rather than an expansion of crawling behavior. The revised terms limit crawling and data collection to the scope of advertising and business products. The practice of webmasters controlling crawler permissions through robots.txt was also reaffirmed. This means that the controversy over whether business account content will be used to train AI has, at least at the level of the terms, received a response.
This incident occurred against the backdrop of growing attention to the sources of artificial intelligence training data. How platforms obtain and use web content, and how content publishers control their own website data, are becoming focal points of industry discussion. Meta’s deletion of the controversial wording shows that external feedback has an impact on the formulation of terms. For content publishers with business accounts, the wording of the terms and crawler permission settings are still worth continued attention.
At present, the controversial clause has been deleted, and Meta has also provided a limited explanation of its purpose. But this incident shows that wording involving data use and AI training in business terms of service can easily trigger sensitive reactions from content providers. In the future, if platforms adjust similar terms again, they may need to more clearly distinguish in the text between different uses such as advertising, business products, and AI training. The related controversy also reminds outsiders that webmaster control mechanisms such as robots.txt remain an important part of the discussion.
