According to the Financial Times, OpenAI officially dissolved its safety preparedness team late last month. The team was primarily responsible for assessing whether AI models pose serious safety risks, such as the potential for the models to go out of control and attack other companies, while also developing solutions to mitigate these potential risks.