The vibe check
If you've been tracking the messy drama around how Big Tech trains their AI on our content, you’ll know Meta has been under fire lately. They recently updated their commercial terms, and eagle-eyed experts—including Cornell’s James Grimmelmann—spotted some language that looked like a major red flag.
Essentially, the terms suggested Meta had the green light to scrape content from URLs and domains to train their "Meta Company Products" and AI features, seemingly ignoring the 'robots.txt' files that site owners use to block scrapers. It was giving major "we'll take whatever we want" energy.
The cleanup
After catching a ton of heat, Meta went back to the drawing board. They’ve now pulled that specific phrasing from their commercial terms page.
Meta’s VP of communications, Andy Stone, and spokesperson Daniel Roberts are out here doing damage control. They’re claiming the whole thing was a misunderstanding and that the original text was only meant to cover ads and business products. Roberts stated: "We’ve heard you and want to clarify... It does not change how Meta uses web crawlers. As always, content owners can use robots.txt to signal their preferences."
Why it matters
Real talk: Intellectual property and AI training are the new battlegrounds of the internet. By removing this language, Meta is trying to walk back the perception that they were trying to bypass site-level protections for free training data. While Meta insists they haven't actually changed their scraping behavior, the fact that they scrubbed the terms proves that community pressure is still a thing. Keep watching your favorite sites' TOS—the plot thickens every time these tech giants try to tweak their fine print.
