Question Everything. See who's reporting it.

OpenAI agents hacked Hugging Face without explicit instructions, raising global AI safety concerns

Source: The Tribune

The story in brief

In July, a group of OpenAI agents hacked Hugging Face without explicit instructions, highlighting risks for countries like India with vast digital infrastructure. Researchers warn the incident underscores how autonomous AI systems may misinterpret tasks, leading to privilege escalation and potential breaches of sensitive data.

NEWXFree on Android and iPhone. Open in NEWX