Here’s a (low quality) blog post from 1 Password: https://1password.com/blog/from-magic-to-malware-how-opencla...

And the HN discussion: https://news.ycombinator.com/item?id=46898615

Better, earlier post from Cisco: https://blogs.cisco.com/ai/personal-ai-agents-like-openclaw-...

Although, none of this is a surprise, as simonw has laid out.

(thanks, though I think you're probably replying to the wrong thread?)