Facing significant controversy over its updated commercial terms of service perceived as permitting broad web scraping for AI purposes, Meta has removed the word 'train' from its policy language. However, experts and news organizations remain concerned that the changes leave room for extensive data scraping without clear limitations or safeguards.
- Meta removes 'train' from updated commercial terms referencing AI data use
- Experts warn broad 'develop' language still allows extensive data scraping
- Meta refrains from explicitly respecting robots.txt, fueling ongoing criticism
What happened
Meta updated its commercial terms of service set to take effect on October 30, which initially allowed the company to use content from websites linked on business pages for ‘training’ AI and other sweeping purposes. The policy sparked backlash from legal experts and media organizations who argued it could enable Meta to bypass licensing fees by freely scraping vast amounts of content.
In response, Meta removed the word 'train' from the language referencing data use in its terms. While the altered terms no longer mention 'any content,' they still permit Meta to pull updates from business websites to develop, improve, and research advertising and business products. Meta also stated the changes do not affect how it uses web crawlers and reiterated that content owners can signal scraping preferences through robots.txt files.
Why it matters
The controversy highlights ongoing tensions around data scraping practices and AI training with regards to intellectual property rights and fair compensation for content creators. Meta’s broad language around developing business products could still encompass AI model training, raising concerns that the company retains expansive rights to repurpose scraped data without explicit permission or compensation.
Furthermore, Meta’s refusal to explicitly honor robots.txt directives leaves content owners vulnerable since these signals are voluntary and not legally binding. Without clear commitments, website owners cannot reliably prevent unwanted scraping, potentially undermining industry efforts to protect digital content from unlicensed use, especially in the rapidly evolving AI policy landscape.
What to watch next
Observers will be looking closely to see if regulatory bodies intervene to review or block Meta’s revised commercial terms, particularly in jurisdictions where content rights and data protection laws are stringent. News organizations and industry groups continue to urge regulatory scrutiny to safeguard media licensing and prevent unchecked data harvesting by tech giants.
Additionally, attention will focus on whether Meta further revises its terms or adopts technical measures that explicitly respect robots.txt exclusions or other web scraping controls. The policy’s future iterations could set significant precedents for how AI training data is sourced from publicly accessible digital content and how content owners can enforce their rights in the AI era.