Meta deletes contentious language from its commercial terms after backlash over AI‑training rights
Meta Platforms Inc. announced the removal of a provision in its commercial terms of use that had sparked criticism from legal observers and publishers. The clause, introduced in a recent update, appeared to give Meta broad permission to harvest content from any website associated with a business account for the purpose of developing and training its artificial‑intelligence (AI) models.
According to TechCrunch, the original wording allowed Meta to “use any content from [a business account’s] domains, web pages, or URLs to provide, develop, improve, research, train, and maintain the integrity of” Meta products and services. Legal experts interviewed by Nieman Lab described the language as a “remarkable provision” because it could be interpreted as granting Meta blanket rights to scrape publisher sites and incorporate that material into its AI training pipelines.
The controversy emerged against a backdrop of ongoing litigation involving AI‑related copyright claims. While Meta successfully defended itself in a high‑profile lawsuit over alleged copyright infringement last year, the company continues to face other legal actions targeting AI developers, including Meta itself.
In response to the criticism, a Meta spokesperson told TechCrunch that the updated terms did not represent a change in how the company crawls the web. The spokesperson emphasized that the clause related only to Meta’s advertising and business products, not to its broader web‑crawling activities.
Subsequently, The Verge reported that Meta had excised the disputed provision entirely. In a statement to the press, Meta clarified:
> “We’ve heard you and want to clarify: the crawling terms we consolidated only applied to our ads and business products, and we’ve clarified that in the terms themselves,” the company said. “It does not change how Meta uses web crawlers.”
According to TechCrunch, the removal signals Meta’s attempt to address concerns from publishers and legal commentators who feared that the company could leverage scraped content for AI training without explicit permission.
Timeline of events
- Initial amendment – Meta revised its commercial terms, inserting language that appeared to grant it sweeping rights to use content from business‑account domains for AI‑related purposes.
- Legal analysis – Experts consulted by Nieman Lab highlighted the potential for the clause to enable large‑scale content scraping, raising red flags about copyright compliance.
- Company response – Meta’s spokesperson asserted that the new wording did not alter the company’s crawling behavior and was limited to advertising‑related services.
- Public backlash – Publishers and advocacy groups expressed alarm, prompting media outlets such as The Verge to investigate the provision.
- Clause removal – Meta announced the deletion of the contentious language and reiterated that its web‑crawling practices remain unchanged.
Contextual background (Opinion)
In my view, the episode illustrates the tension between the rapid expansion of AI capabilities and existing copyright frameworks. By initially embedding a clause that could be read as authorizing unrestricted data harvesting, Meta exposed itself to scrutiny from both the legal community and content creators who rely on the protection of their work. The swift removal of the language suggests that the company is responsive to stakeholder pressure, yet the underlying question—how tech firms will legally source training data for AI—remains unresolved.
Potential implications
According to TechCrunch, the clarification may ease immediate concerns among publishers who feared that Meta could repurpose their articles for AI models without consent. However, the statement that the change “does not change how Meta uses web crawlers” leaves open the possibility that the company will continue to collect publicly available data for purposes unrelated to its advertising and business products.
Legal analysts note that the episode could set a precedent for how other technology firms draft and revise their terms of service, especially as regulators worldwide examine AI training data practices. The removal also underscores the influence of public and industry feedback on corporate policy, demonstrating that even large platforms must navigate the expectations of content owners and legal experts.
Looking ahead
The removal of the clause does not, according to Meta’s statement, alter the company’s overall web‑crawling methodology. As AI models become increasingly central to product development, the balance between data acquisition and intellectual‑property rights will likely continue to generate debate. Stakeholders will be watching future updates to Meta’s terms for any signs of renewed attempts to broaden data‑use permissions.
Conclusion
Meta’s decision to delete the controversial provision from its commercial terms reflects a response to legal criticism and publisher concerns over AI training data. While the company maintains that its web‑crawling practices remain unchanged, the episode highlights the broader industry challenge of aligning AI development with copyright law. Ongoing scrutiny from regulators, legal experts, and content creators will shape how Meta and similar platforms negotiate the use of publicly available content in the AI era.
Source: TechCrunch
Frequently Asked Questions
What was the controversial provision in Meta’s commercial terms?
The provision allowed Meta to “use any content from [a business account’s] domains, web pages, or URLs to provide, develop, improve, research, train, and maintain the integrity of” its products and services.
Why did legal experts consider the clause “remarkable”?
They believed it could give Meta permission to scrape any publisher’s website linked to a business account and use that content to train its AI models.
Has Meta changed its overall web‑crawling practices?
According to Meta’s statement reported by TechCrunch, the removal of the clause does not change how the company uses web crawlers.