OpenAI promises that its future models won't be as censored & will allow adult content

OpenAI has just recently updated its guidelines and promised that future AI models will offer more freedom when handling controversial topics.

The company’s expanded 63-page Model Spec sets clear rules that ensure its models do not avoid or limit discussions on sensitive subjects. OpenAI says this update is designed to increase transparency and allow users more control over how the AI responds, promoting a spirit of intellectual freedom.

The new guidelines say that the models should not steer users toward any particular agenda, no matter how subtle.

By doing so, OpenAI aims to prevent any bias from shutting out certain viewpoints, a move that comes as critics—including high-profile figures like Elon Musk and David Sacks—have accused current models of being overly cautious or “woke.”

The Model Spec also tackles how the AI should approach complex and controversial scenarios, such as ethical dilemmas and historically charged topics.

The update also acknowledges feedback from users requesting a “grown-up mode” for mature content, and it’ll allow certain adult content like erotica in appropriate contexts although revenge porn and deepfakes are still strictly banned.

In an interview with The Verge leading up to the release of the updated guidelines, OpenAI’s Joanne Jang said, “We can’t create one model with the exact same set of behavior standards that everyone in the world will love.”

The Microsoft-backed company has recently removed its ChatGPT model picker by merging the o-series and GPT-series models into one unified system. OpenAI’s GPT-4.5 and GPT-5 are also expected to launch in the next “weeks or months,” according to CEO Sam Altman.