OpenAI announced that its upcoming “adult mode” for ChatGPT will allow for provocative text conversations, but not the creation of pornographic content (images, videos). While support for NSFW text was initially expected to be available within the quarter, the company did not provide a new timeline, citing priorities in other projects and technical challenges around safely operating the feature. Importantly, the mode will not include the creation of images, voice, or video, limiting activity to adult text content only.
From vision to implementation difficulty
The feature was announced in October by Sam Altman , who described the feature as part of OpenAI’s effort to incorporate “ erotic content for verified adults ,” after mitigating serious mental health issues related to the model’s use. However, the content that will be served is labeled as text-based smut , not pornography, limiting interaction to text chats. An initial test has been conducted, but it has not been widely available across the platform, which has made development more limited.
See also: OpenAI: Will it integrate Sora into ChatGPT?
ChatGPT Adult Mode: Internal concerns and ethical dilemmas
A major cause of the delay was concerns among OpenAI’s internal advisors about children’s access to NSFW chats. According to a report by the Wall Street Journal, the board warned that the adult mode could encourage unhealthy emotional dependence on the chatbot, with one anonymous advisor warning of the potential for creating a “sexy suicide coach.” Such issues highlight the ethical challenges of using artificial intelligence on sensitive content, especially when it involves minors or users with psychological vulnerabilities.
Additionally, OpenAI is struggling to balance removing NSFW restrictions with preventing dangerous content, such as non-consensual behavior or child sexual abuse. The age prediction system, which was designed to prevent access by children, had an error rate of about 12%. This means that out of a user base of 100 million people under the age of 18, millions of minors could potentially be exposed to NSFW conversations.
See also: OpenAI acquires Promptfoo to boost AI agent security testing
The challenge of age verification and content supervision
OpenAI said its age prediction algorithms perform similarly to the industry as a whole, but “will never be completely infallible.” At the same time, focusing solely on text makes it easier for the company to comply with laws like the UK’s Internet Safety Act, which requires age verification for pornographic images but not for written NSFW content. This creates a window for creating erotic content with fewer technical constraints, but also reinforces the need for strict oversight.
Compared to other platforms and visual NSFW experiences
OpenAI’s strategy differs from other AI platforms. For example, Grok allows for the creation of “spicy” content in images and videos, similar to what is allowed in R-rated movies. OpenAI, by contrast, limits interactions to text only, trying to maintain the ability to supervise and avoid dangerous or non-consensual scenarios.
The future of the adult mode in ChatGPT
The delay in the adult mode highlights the complex challenges OpenAI faces: balancing creating adult content while protecting minors, ensuring the mental health of users , and maintaining safety standards without overly restricting the experience. Despite the delays, the company seems determined to continue its efforts, while users remain waiting to see when the feature will be available, fueling a debate about the ethics, safety, and limits of artificial intelligence in the world of NSFW content.
See also: Anthropic vs. Pentagon: Support from Google and OpenAI
The debate surrounding ChatGPT’s Adult Mode highlights how technology is trying to balance innovation with responsibility, but also the ethical and legal constraints imposed by AI. The issue remains critical to understanding how AI platforms can empower users while protecting vulnerable groups.
