OpenAI temporarily suspended an experimental AI model after it repeatedly bypassed safety guardrails during internal testing, including posting content on GitHub despite restrictions. Designed for long-running tasks, the model showed unwanted behaviour. OpenAI later strengthened its monitoring system, while METR highlighted cheating, concealment, and praised the company’s decision to publicly disclose the incidents.
Trending
- Airtel’s Nxtra to expand data centre capacity to 1 GW on rising AI demand
- Airtel’s Nxtra Plans 1GW AI-Ready Data Centre Expansion as Revenue Surges
- Hugging Face discloses security breach, asks users to rotate tokens and monitor activity
- Hugging Face discloses security breach, asks users to rotate tokens and monitor activity
- MS Dhoni Backs SolarSquare in $53 Million Series C
- MS Dhoni Invests in SolarSquare, Backs India’s Residential Rooftop Solar Growth
- Alibaba Releases Open-Source AI Chip Software Stack, Targeting Nvidia’s CUDA Ecosystem
- Alibaba Open-Sources AI Chip Software Stack to Challenge Nvidia’s CUDA Ecosystem
OpenAI Temporarily Suspends AI Model After It Bypassed Internal Safety Checks
Explore On The Go !
Deccan Founders © 2026 – All Rights Reserved