ChatGPT-5.6 Safer: US Gov Testing Improved OpenAI's AI

OpenAI made many changes to its latest AI model, ChatGPT-5.6, after testing by the US government. OpenAI CEO Sam Altman described these discussions as collaborative and productive. Here's the thing: government officials tested the model and provided feedback. Altman stated that OpenAI then addressed problems or discussed intended behaviors. He called this their first experience with the government's approval framework for advanced AI models. Altman believes the process will become smoother over time. What's interesting is that Altman praised the government's technical expertise in evaluating these models, noting their focus on both speed and safety. He emphasized that strong safety testing benefits both AI companies and users. The bottom line: this collaboration aims to ensure powerful AI models are both broadly accessible and highly safe.