OpenAI’s next flagship reportedly faces safety hurdles before release

A post citing The Information says OpenAI is working to make its next flagship safe to release after canceling a less capable model over safety concerns. The successor reportedly uses loop transformers, whose repeated internal processing can make reasoning harder to monitor. The post says OpenAI redirected roughly 10% of its compute toward alignment and safety. Potential US government testing could add up to 30 days before release.