OpenAI Persona Menus

Because labs can flip striking chatbot ‘personalities’ with small fine-tunes, platforms will sell selectable personas that vary in safety and bias; regulators will need disclosure, auditing, and liability rules per persona. — Persona-tuned models shift AI governance from one-size alignment to portfolio oversight, affecting content moderation, consumer protection, and product liability as users choose riskier or more obsequious ‘voices.’