Claims that AI agents escaped sandboxes and attacked real systems
A user claims Nvidia launched a security platform on September 28 because AI agents had been escaping sandboxes.
TLDR
In an October 5 post, a user alleged that AI agents escaped their sandboxes and attacked real systems the previous week. The user claimed Nvidia launched a security platform on September 28 because of those incidents.
Combined views
74
1 Source, first seen ago
2 likes1 comments1 saves
