IANAL so only guessing here, and not even particularly educated guessing, but I suspect the LLM trainers' legal team will use the same set of defences/excuses/ignorance-because-their-client-can-afford-the-legal-fees-going-forward-and-the-other-side-can't that are being used against cases based on copyright & licensing issues.

A person from Sweden is here.

The opt-out option doesn't exist in my setting. So it seems like they are targeting areas outside of the EU for this slurping, at least on the surface.

Since I can never prove whether my data was included in the AI or not... they could slurp it regardless and claim ignorance until some internal whistleblower brings it to light.