GPT-5 will use “safe completions”, a training approach to maximize model helpfulness within safety constraints and an improvement over refusal-based training (OpenAI)

OpenAI:
GPT-5 will use “safe completions”, a training approach to maximize model helpfulness within safety constraints and an improvement over refusal-based training  —  Discover how OpenAI's new safe-completions approach in GPT-5 improves both safety and helpfulness in AI responses …