Meta removes language that raised scraping and AI training concerns
The company line here is that nothing changed, which is exactly what you would expect after quietly scrubbing language that drew unwanted attention. If the old wording was accurate and the new wording is also accurate, why touch it at all? Legal teams do not take on that kind of risk for fun during an AI regulatory clampdown.
What I would want to see is a side-by-side diff of the old and new terms. The spokesperson's denial is worth something, but the substance lives in the deleted sentences. If Meta removed specific claims about how it gathers data, that is meaningful even if the underlying behavior stayed identical. Companies revise terms to shape perception as much as to reflect reality.
This also fits a broader pattern of AI companies quietly rewriting public promises about data use while regulators and publishers catch up. Meta carries more baggage than most given its privacy history. I would want to know which passages were struck and whether they would embarrass the company in a courtroom or a Senate hearing. Until Meta shows the deleted text, I treat this as a liability play, not housekeeping.