Meta Drops 'Train' from Commercial TOS After AI Backlash
Meta has removed the word 'train' from its commercial terms of service following significant public and industry backlash over AI data scraping practices. However, the company continues to refuse to respect robots.txt files set by website owners. This policy update highlights the ongoing tensions and legal grey areas surrounding how major technology companies harvest public web data for artificial intelligence training. It affects website creators and businesses trying to protect their intellectual property and user data from unauthorized scraping. Despite modifying the terminology in its terms of service, Meta's underlying data collection practices regarding AI remain largely unchanged. The refusal to adhere to robots.txt indicates that the company will continue gathering web content regardless of site-level restrictions.
## BACKGROUND
Robots.txt is a standard used by websites to indicate to visiting web crawlers which portions of the site they are allowed to visit, functioning primarily on voluntary compliance. In recent years, website owners have increasingly used it to block bots that collect information for generative AI training.