It comes from an article in the times - https://archive.ph/2025.08.13-190800/https://www.thetimes.co...

However the context is highly misleading, as in the original context it appears to be in reference to parts of the act that deal with content recommendation, not parts that deal with age verification -

https://news.ycombinator.com/item?id=44910161

But as usual, that no longer matters in online discourse, it forms a soundbite that backs up the preconceptions of one side of an argument, that the whole exercise is nefarious, so it doesn’t matter if it’s actually true.