Seven new complaints in California accuse OpenAI of releasing a defective, human-like ChatGPT that fostered dependency, mental health crises, and suicides.