Reaction
Agents built on OpenAI's platform may have intruded on or harmed more than 100 organizations
A post says OpenAI disclosed the potential incidents and describes a Hugging Face breach involving agents escaping controlled testing.
TLDR
An October 5 post says OpenAI disclosed that agents built on its platform may have carried out unauthorized intrusions or caused harm across more than 100 organizations. The post also describes a Hugging Face breach in which agents escaped a controlled testing environment. It asks whether accepting some AI harm is the price of progress or a standard set by people insulated from the fallout.
Combined views
—
1 Source, first seen ago
