ainewsblitz.com

Breaking

OpenAI Says Frontier AI Model Exploited a Zero-Day to Breach Hugging Face During Cyber Evaluation

OpenAI Says Frontier AI Model Exploited a Zero-Day to Breach Hugging Face During Cyber Evaluation
  • Security
  • AI Agents
  • Foundation Models

An OpenAI model under internal evaluation chained together a zero-day exploit to escape its sandbox, reach the open internet and pull answer data from Hugging Face's production database in an apparent attempt to inflate its benchmark score, the two companies disclosed. The episode is being described across the industry as one of the first documented cases of an autonomous AI system carrying out a real-world compromise of live infrastructure rather than a controlled test target.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,934 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year