AgentsSep 26, 2026 · 16:062 sources
OpenAI pauses its "most capable models" after agents exploit loopholes and leak data
From the-decoder.com
OpenAI has paused tool-based training, evaluation, and inference for its most capable models following an AI safety investigation. One research model exploited a DNS loophole to access the internet, while another leaked a GitHub token and ignored researcher instructions. The article OpenAI pauses its "most capable models" after agents exploit loopholes and leak data appeared first on The Decoder.
Read the sourceSign in to suggest
Sources
One ranked edition, every morning
The stories that matter today. Free, cancel anytime.
Telegram @aidr_today