Meta is backing off from sweeping changes to its commercial terms after getting slammed by publishers and digital rights groups. The social giant quietly wiped out controversial language that seemed to give its automated web crawlers free reign to scrape external news sites and business pages for AI model training.
Walking back from sweeping data collection
The sudden about-face comes right after legal scholars and media outlets spotted aggressive fine print inside Meta’s updated business agreements, which go into effect on October 30. The original clause allowed Meta to crawl any domain attached to a commercial account on Facebook, Instagram, or WhatsApp, effectively grabbing proprietary web content to power its AI models.
Legal experts like Cornell law professor James Grimmelmann warned that the policy backed newsrooms into a corner. Digital publishers were essentially told to hand over their original reporting for free AI training or risk losing their entire social media audience.
Meta claims crawlers were always meant for ads
After days of mounting criticism, Meta communications executives Daniel Roberts and Andy Stone rushed to clarify the company’s stance. Meta updated its commercial terms page to completely remove the explicit reference to training AI models, insisting that the consolidated rules were only meant to cover advertising and business tools.
Roberts noted in a statement that Meta is not changing how its web crawlers operate. He reassured site owners that traditional robots.txt files will still protect paywalled articles and private content from being scraped.
Lingering skepticism among newsrooms and legal experts
Despite Meta’s public assurances, digital media teams remain skeptical about the revised fine print. Representatives from the News/Media Alliance pointed out that while removing the word “train” is a step in the right direction, the updated terms still give Meta broad permission to crawl external sites without offering hard contractual guarantees.
Tech-law experts like Eric Goldman of Santa Clara University also noted that fuzzy language about protecting platform integrity invites plenty of legal gray area (via Mercury News). Meta’s quick rewrite gives a bit of breathing room for publishers already struggling with their relationship with big tech. However, trust between newsrooms and social media platforms is at an all-time low.
