OpenAI has launched a new large language model, GPT-5, but it may not have ruffled too many feathers this time ...

Study reveals ChatGPT's potentially dangerous responses to prompts about self-harm, eating disorders, and substance abuse, raising concerns.