OpenAI delays ‘adult mode’ for ChatGPT to focus on work of higher priority | OpenAI | The Guardian

OpenAI has officially shelved plans to introduce a user-controlled 'NSFW' toggle for its flagship ChatGPT model, a feature initially floated in a May research paper. The company confirmed to The Guardian that it is no longer actively developing this capability, stating it is instead focusing its safety efforts on 'other safety work of higher priority.' This represents a significant strategic pivot, moving away from a model that would have allowed users to opt into generating sexually explicit or violent content within defined boundaries.

The proposed feature, often colloquially termed 'adult mode,' sparked immediate controversy and debate within and outside the company. Critics argued it could undermine OpenAI's safety-first brand and be misused to generate harmful material, despite proposed guardrails. The decision to delay it indefinitely suggests OpenAI is prioritizing the development of core AI capabilities, enterprise tools, and foundational safety research over exploring more permissive content-generation frontiers. This move aligns with a broader industry trend of tightening content moderation as regulatory scrutiny intensifies globally.