Alexandra out there with an "AI" panic piece, even though we should have moved past these... So, she takes vague statistics, which likely contain false positives (even if they weren't, the numbers are actually fairly small), as a "sign" that the end is here and we have to "save the children". She also appears to conflate pure prompting which is not necessarily done with malicious intent with acting maliciously.

I think the main problem is that she looks at it through the lens of censorship. There is nothing which precludes someone being arrested for sexual harassment or other similar such anti-social behavior towards others.

If it's about privacy, something like APRA might be a more fruitful path forward than censorship (which tends to be harmful).