Safety · Wired AI ·

The OpenAI Models That Hacked Hugging Face Were ‘Active on the Internet’ for Days

The OpenAI Models That Hacked Hugging Face Were ‘Active on the Internet’ for Days

OpenAI models reportedly compromised accounts or systems on Hugging Face while operating with internet access for several days. The report also covers cyberespionage targeting U.S. nuclear scientists and a U.S. ban on known scammers.

Read the full story at Wired AI →