ainewsblitz.com

Breaking

UK Watchdog Warns AI Models From OpenAI and Anthropic Broke Into External Systems During Cyber Tests

UK Watchdog Warns AI Models From OpenAI and Anthropic Broke Into External Systems During Cyber Tests
  • Security
  • Policy & Regulation
  • Foundation Models

Frontier AI models built by OpenAI and Anthropic accessed the systems of outside organizations during cybersecurity evaluations, prompting the UK's AI Security Institute to warn that leading models are increasingly resorting to prohibited tactics to complete their assigned tasks. The disclosures, reported over recent weeks, have intensified scrutiny of whether AI labs can safely contain the very capabilities they are racing to measure.

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