OpenAI-Mitarbeiter haben vor Monaten wegen Schießerei in Kanada Alarm geschlagen | Der ChatGPT-Hersteller hat sich im vergangenen Juni dagegen entschieden, die Behörden über die Gewaltbeschreibungen von Jesse Van Rootselaar zu informieren.

I understand the need to flag and review these cases, potentially report to police when there is credible evidence of harm to self or others. 

But something about how easily they can access chat bot conversations makes me uneasy, especially given the amount of confidential information people are giving them. 

It is very disappointing to read this. The automated system did its jobs, and the humans said nothing. What would’ve been wrong with sending this chat to the local police?

If you openly wonder about killing people, and ask how to do it, do you really have a right to privacy?

Can not agree with „Private Chat can be informed by Authorities“, it is gonna a be a ultra slippery slop. The mentality of „Sacrificing your rights/privacy to maybe save lives“ had shown again and again that it prevent nothing but general election wedge issues and papaer criminals.

Call me a conspiracy theorist but I’m pretty suspicious of an article coming from the WSJ that is basically promoting OpenAI monitoring people’s behaviours. I’m referring to the Murdoch family signing a licensing agreement with OpenAI in 2024. Just saying.

Paywalled