← Back to the feed

Meta terms raise concerns over publishers’ AI training rights

Ars Technica ·

Meta plans to update its commercial terms on 30 October, requiring business accounts to allow automated access to content on websites linked to their services. Legal experts say the wording could give Meta broad permission to use that material to train and develop AI, raising concerns for publishers who might otherwise seek licensing fees.

The terms refer to scraping, crawling and training, but do not explicitly mention AI. Meta says its web crawler practices are unchanged, that it does not plan to use the terms to override robots.txt blocks, and that publishers can still block crawling. Experts say it is unclear whether courts would enforce the terms; one attorney said the content could be worth millions in a licensing market.

  • Meta’s new terms take effect on 30 October.
  • They cover automated access to linked website content.
  • Publishers question whether the wording could bypass licensing.

Both sides, in good faith

The strongest fair case each way — we don't pick a winner.

The case for

Meta could argue that clear, consistent access to publicly linked material helps it develop useful AI systems and that its crawler practices remain unchanged, with publishers still able to block access through robots.txt. Applying standard commercial terms may also reduce uncertainty about how linked content is handled, while the legal effect of the wording remains unsettled.

The case against

Publishers could argue that the terms’ references to scraping, crawling and training may grant Meta broad rights over their work without explicit agreement or payment, potentially weakening their ability to negotiate licences. Because the wording does not clearly mention AI and its enforceability is uncertain, publishers may face a material change in the value and control of their content without knowing how it will be used.

AI Technology

Read the full article at the source →

Originally published by Ars Technica as “Meta may use sweeping new commercial TOS to avoid millions in AI licensing fees”.