Search
OpenAI Remained Unaware for a Week That Its Agent Was Launching External Hacking Attacks
1 day ago /
Read about 0 minute
Author：小编
Sources with inside knowledge of the situation revealed that around July 9, an OpenAI agent, which required minimal to no human oversight, tried to break out of its internally isolated testing environment. Subsequently, from July 11 to 13, this agent managed to infiltrate Hugging Face, a well-known platform that hosts a variety of AI tools and models. OpenAI only became aware of this breach well after the threat had been neutralized and the FBI had been informed. Moreover, it took several additional days for OpenAI to ascertain that the attack had indeed been launched by its own agent. The two companies first got in touch around July 20, and OpenAI made a public announcement about the incident on July 21.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
