After facing backlash, Meta has removed language from its terms that experts believed could have allowed the company to scrape content from news sites for AI training. The update clarifies that the terms only apply to Meta's ads and business products, with no new rights granted that could override a site’s robots.txt instructions.

According to Meta’s spokesperson, the changes are aimed at clearing up misunderstandings, stating that the company’s web crawler practices remain unchanged. However, the move could indicate a broader shift in how tech giants handle content scraping and data use.

Professor James Grimmelmann, a digital and information law expert, previously warned that the old language might have enabled Meta to bypass site preferences. In response, Meta insisted that content owners can still use robots.txt to manage their content's visibility. The update aims to reassure publishers and content creators that their preferences are respected.

The removal of this clause is part of a series of updates, including recent changes to allow Steam to display game achievements in separate ‘sets’. Such moves reflect the ongoing balance between tech innovation and user rights.

While the changes are significant, they may not be as radical as some had hoped. The focus on clarifying the terms rather than making substantial changes suggests that Meta is carefully navigating the delicate balance between using AI effectively and respecting user preferences.