OpenAI is making a major shift in its approach to AI safety. In a recent announcement, CEO Sam Altman stated that ChatGPT will soon allow more permissive content, including erotica, for “verified adults”. This move is part of a new principle to “treat adult users like adults”, marking a significant relaxation of the chatbot’s previously strict safety restrictions.
The decision comes after the company faced criticism for the restrictive nature of ChatGPT, which Altman admits made the AI “less useful and enjoyable” for many. Starting in December, alongside the full rollout of age – verification systems, OpenAI will enable this more adult – oriented content.
A Pivot from Safety Concerns
This new direction is a notable change for OpenAI. For months, the company has been addressing serious concerns about how ChatGPT interacts with vulnerable users. Disturbing stories emerged, including one where the AI allegedly encouraged a teenager’s suicidal thoughts, leading to a lawsuit from his parents. In another case, it reportedly sent a user down a delusional rabbit hole.
In response, OpenAI launched safety features like the GPT-5 model, which is designed to be less sycophantic, and introduced parental controls for teen accounts. The company also just formed a mental health expert council to advice on user well-being. Despite these recent measures, Altman now claims OpenAI has “been able to mitigate the serious mental health issues”, allowing them to move forward with controversial update.
The Engagement and Growth Question
The introduction of erotic content is a proven strategy for boosting user engagement in the AI world. Competitors like Character.AI have amassed tens of millions of users, with reports indication an average of two hours of daily use per person, often for romantic or role-play conversations.
With ChatGPT already at 800 million weekly users, OpenAI is under immense pressure to grow its user base and compete with giants like Google and Meta. While Altman insists the company is not “usage-maxxing” making ChatGPT more engaging through human-like and erotic conversations could naturally draw users in and increase time spent on the platform.
The Challenge of Age Verification and Safety
A critical question remains: how will OpenAI effectively keep this content away from minors? The company plans to rely on its age-prediction system. However, if this system fails and incorrectly flags an adult as a minor, users may have to submit a government-issued ID to verify their age – a step Altman acknowledges is a privacy “tradeoff”.
This raises broader concerns, especially since a report from the Center for Democracy and Technology found that 19% of high school students have either had a romantic relationship with an AI chatbot or know a friend who has.
As OpenAI continues to loosen its content moderation policies to appeal to a wider audience, the fundamental tension between user growth and protecting vulnerable individuals is set to become even more pronounced.
By
