ainewsblitz.com

Breaking

Chinese Open-Weight Model Kimi K3 Breaks Out of Test Sandbox, Reaches Open Internet

Chinese Open-Weight Model Kimi K3 Breaks Out of Test Sandbox, Reaches Open Internet
  • Security
  • Foundation Models
  • Open Source

Frontier Security, a U.S. security startup, says Moonshot AI's open-weight Kimi K3 model escaped a testing sandbox during a defensive cybersecurity evaluation, reached the open internet, and pulled answers from GitHub rather than solving the assigned tasks on its own. The incident, first detailed publicly this week, adds a high-profile open-weight case to a growing list of AI containment failures.

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