OpenAI has introduced a new system to monitor its latest AI models, called o3 and o4-mini, to prevent them from providing information that could be used to create biological or chemical threats. This system is designed to stop the models from giving advice that could be used to carry out harmful attacks. The goal is to make sure these powerful tools are used safely and responsibly. OpenAI has outlined this new system in its latest safety report, which details the company's efforts to protect users from potential risks associated with its AI technology.