The laws aren't 100% settled, the latest reveals of MS execs in the New York Times lawsuit (or was it the New Yorker?) that came up here the other day was quite damning.

Right now, only that lawsuit seriously challanges big-AI, but _if_ they succeed then everyone who's had something ingested has a precedent to lean on for a lawsuit against the big-AI corps (and potentially by extension their users).

So more worryingly for free software is that snippets that has found their way via LLM models into openly licensed projects, this could be commercial "shared source" projects that have gotten ingested or even internal tools since the AI corporations trains their models on your prompts, projects and even git histories.