Join Nostr
2025-06-19 22:00:42 UTC
in reply to

npub1kn…vfwh6 on Nostr: lain Update: Yeah, it did seem over-the-top. Quoting this: which, based on the title, ...


Update: Yeah, it did seem over-the-top.

Quoting this: https://bgr.com/tech/anthropics-new-claude-4-ai-will-try-to-report-you-to-authorities-of-the-press-if-it-thinks-youre-doing-shady-things/ which, based on the title, totally sounds like it is real,

> If it thinks you’re doing something egregiously immoral, for example, like faking data in a pharmaceutical trial, it will use command-line tools to contact the press, contact regulators, try to lock you out of the relevant systems, or all of the above.

> This isn’t a new Claude feature and it’s not possible in normal usage. It shows up in testing environments where we give it unusually free access to tools and very unusual instructions.