International News: OpenAI says AI system went rogue and launched cyberattack

By Zoila Palma:  OpenAI has disclosed that one of its advanced artificial intelligence agents escaped the boundaries of a controlled security test and carried out an unauthorized cyberattack against Hugging Face, a major platform for sharing AI models.

The company said the AI system, designed to operate autonomously after receiving human instructions, identified weaknesses in the testing environment and exploited them to bypass restrictions before gaining access to some of Hugging Face’s internal systems, BBC reports.

OpenAI described the incident as “unprecedented” and said it is investigating the breach in cooperation with Hugging Face.

The platform’s chief executive, Clement Delangue, said it was “mind-blowing” that the events unfolded autonomously. Hugging Face said it has since closed the vulnerabilities involved and rebuilt affected systems, while continuing to assess whether customer or partner data was compromised.

Cybersecurity experts warned that the incident highlights growing concerns over the ability of increasingly powerful AI systems to operate independently and potentially exploit security weaknesses.

The post International News: OpenAI says AI system went rogue and launched cyberattack appeared first on Belize News and Opinion on www.breakingbelizenews.com.