Meta removes training language from its website-crawling terms
Meta removed the disputed training language from its commercial terms after criticism of the permission it appeared to grant over website content. The company says the wording concerns advertising and business products and leaves crawling practices unchanged. Its new statement also says site owners’ robots.txt preferences continue to apply. The change is in the contractual wording, following the earlier dispute over its scope.
Artificial Intelligence··Morning
The disputed training sentence is removed
Meta removed explicit training language from the website-crawling section of its commercial terms. The company’s revised wording limits the stated purposes to advertising and business products. The change follows objections that the earlier provision gave broad permission to use content from websites associated with business accounts. Those accounts operate across Meta services including Facebook, Instagram and WhatsApp. The new terms are scheduled to take effect on October 30.[1], [2]
Earlier wording drew objections over publisher content
The earlier list of purposes included providing, developing, improving, researching and training Meta products. James Grimmelmann, a digital-law professor at Cornell University, was among the experts who interpreted it as permission that could reach news websites and AI training. Reporting by journalism research publication Nieman Lab brought those legal readings to attention. A central concern was whether contractual permission might override preferences stated in robots.txt, the file used to signal access preferences to automated crawlers.[1]
Meta maintains that its crawler practices are unchanged
Meta communications executive Andy Stone and spokesperson Daniel Roberts had disputed the broad interpretation. In a new statement on Saturday, Roberts said the consolidated crawling terms concerned advertising and business products. He also said content owners could continue to signal their preferences through robots.txt and that the update gave Meta no new rights overriding those preferences. The removed language changes the contract’s wording; the company continues to deny changing how it uses web crawlers.[1]
