AI Safety

OpenAI has officially launched a new misalignment reporting framework, accompanied by six detailed incident reports, according to a recent report...