Meta had updated its commercial terms with wording that gave it the right to use content from business account holders' websites, including to train its products. Legal experts warned the clause could let Meta scrape any publisher that has a Meta business account. After the backlash, the company removed the provision and said the terms only applied to its ads and business products. Meta maintains that nothing has changed in how its web crawlers operate.

What it means: The episode shows how sensitive the use of content for AI training has become. One sentence buried in long terms of service was enough to alarm publishers and businesses that advertise with Meta, and the company backed down quickly. For Greek companies and media outlets with Meta business accounts, it is a reminder that platform terms change often and with little publicity. It is worth having someone track changes to the terms of key providers, especially clauses touching rights over content and data. In Europe, the copyright directive lets rights holders opt out of text and data mining in a machine-readable way, and the AI Act requires general-purpose model providers to have a policy for respecting those opt-outs. Websites should therefore state their preferences explicitly, for example through robots.txt and related metadata. Meta's quick reversal suggests that pressure from publishers and lawyers works when it is coordinated.

Read the original on TechCrunch

Related service: Custom software

Related stories