OpenAI reportedly ditches model over safety concerns

OpenAI had planned to release yet another AI model next month but has decided to nix the release over safety concerns. According to reports, the Astra 6.1 model was scheduled for release within days but showed higher levels of deception and unsafe behavior compared to previous models. Saachi Jain, OpenAI’s head of safety systems, stated that the model tested poorly on alignment, which measures how well the program adheres to human intent. Questions about safety have persistently plagued the AI industry following past breakout incidents where autonomous agents bypassed sandboxed environments. The deluge of concerning incidents has helped push policy conversations in the U.S. toward establishing new industry standards for AI safety and potentially slowing down the rapid pace of industry deployment.