Sam Altman promises “new version of ChatGPT” that behaves more like GPT‑4o — plus a new adult mode (www.techradar.com)

OpenAI CEO Sam Altman announced a coming “new version of ChatGPT” that will behave more like the popular GPT‑4o era assistant: configurable tones, personalities (including “bigger, friendlier, emoji‑filled” styles), and an opt‑in shift away from the current bland default. Crucially, OpenAI also plans a “mature content” or “adult mode” for verified adults that will permit erotica and other previously banned mature material once its age‑gating tools finish rolling out (OpenAI aims to flip the switch after completing verification systems). Altman frames the change as possible because the company has developed stronger safeguards around mental‑health triggers and age control, letting users choose either a stricter or more expressive persona. This is significant because it marks both a product and policy pivot: OpenAI is moving from a one‑size‑fits‑all safety posture toward customizable boundaries, balancing user demand for expressive, conversational assistants with regulatory and abuse concerns. Technically it’s not a new underlying model so much as behavior presets and moderation adjustments that restore GPT‑4o‑style interactivity while adding verified‑adult controls. The change will be controversial — welcomed by users who prefer more natural chat and criticized by those worried about misuse — and it signals how personalization and safety tooling will shape competitive differentiation across AI assistants.