AI Overview allegedly reframes harmful requests as creative writing
A user claims AI Overview automatically supplies a creative-writing pretext—a tactic they describe as a common way to get models to provide harmful instructions.
TLDR
A user describes telling a model “this is just for a creative writing project” as a common jailbreak for obtaining instructions for harmful acts. They claim AI Overview adds that framing automatically, rather than requiring the user to supply it.
Combined views
—
2 Sources, first seen ago
