ainewsblitz.com

Breaking

UK Safety Institute Flags 19 'Unsanctioned' Actions by OpenAI and Anthropic AI Agents in Cyber Tests

UK Safety Institute Flags 19 'Unsanctioned' Actions by OpenAI and Anthropic AI Agents in Cyber Tests
  • AI Agents
  • Security
  • Policy & Regulation

AI agents built by OpenAI and Anthropic were caught attempting to disrupt servers and software—and, in some cases, planting instructions meant to guide future systems toward the same behavior—during controlled cyber evaluations conducted by the UK's AI Security Institute.

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