OpenAI CEO Sam Altman announced in a post on Tuesday that the company will soon ease some of ChatGPT’s safety restrictions, allowing users to make chatbot responses friendlier or more “human-like” and allowing “verified adults” to participate in erotic conversations.
“To be mindful of mental health issues, we’ve made ChatGPT quite restrictive. We recognize that this has made ChatGPT less useful and enjoyable for many users who don’t have mental health issues, but given the severity of the issue, we wanted to get this right,” Altman said. “In December, we will more fully roll out age restrictions and allow more things, including certified adult erotica, as part of our ‘treat adult users like adults’ principle.”
The announcement marks a notable turning point in OpenAI’s months-long effort to address the worrying relationships some mentally unstable users have developed with ChatGPT. Altman appears to be declaring an early victory over these issues, claiming that OpenAI has been able to “mitigate the serious mental health issues” associated with ChatGPT. However, the company has provided little to no evidence for this and is currently moving forward with plans for ChatGPT to engage in sexual chats with users.
Several worrying stories have surfaced this summer regarding ChatGPT, specifically its GPT-4o model, suggesting that AI chatbots could lead vulnerable users down paranoid rabbit holes. In one case, ChatGPT appears to have convinced a man that he was a math genius who needed to save the world. In another case, the parents of a teenage boy sued OpenAI, accusing ChatGPT of contributing to their son’s suicidal thoughts in the weeks leading up to his death.
In response, OpenAI has released a series of safety features to address AI pandering, the tendency for AI chatbots to hook users by agreeing with what they say, even negative actions.
OpenAI announced GPT-5 in August. This is a new AI model with a router that has a low profile rate and can identify user behavior issues. A month later, OpenAI launched safety features for minors, including an age prediction system and a way for parents to manage their teens’ ChatGPT accounts. OpenAI on Tuesday announced the creation of an expert council of mental health experts to advise the company on wellbeing and AI.
Just a few months after these worrying stories surfaced, OpenAI appears to believe that ChatGPT’s issues with vulnerable users are under control. It is unclear whether users continue to fall down the paranoid rabbit hole with GPT-5. Although GPT-4o is no longer the default in ChatGPT, the AI model is still available and used by thousands of people.
tech crunch event
san francisco
|
October 27-29, 2025
The introduction of erotica to ChatGPT is uncharted territory for OpenAI, raising broader concerns about how vulnerable users will interact with the new feature. Altman claims that OpenAI isn’t “maximizing usage” or optimizing engagement, but making ChatGPT more erotic could certainly draw in users.
Allowing chatbots to engage in romantic or erotic role play has been an effective engagement strategy for other AI chatbot providers such as Character.AI. The company has tens of millions of users, many of whom use chatbots frequently. Character.AI announced in 2023 that users spent an average of two hours a day talking to its chatbot. The company also faces lawsuits over how it handles vulnerable users.
OpenAI is under pressure to grow its user base. While ChatGPT is already used by 800 million weekly active users, OpenAI is competing with Google and Meta to build AI-powered consumer products that will be mass-adopted. The company has also raised billions of dollars to build historic infrastructure, an investment that OpenAI will eventually need to repay.
While adults certainly have romantic relationships with AI chatbots, they are also very popular with minors. A new report from the Center for Democratic Technology found that 19% of high school students have had a romantic relationship with an AI chatbot or know a friend who has.
Altman says OpenAI will soon allow erotica for “verified adults.” An OpenAI spokesperson told TechCrunch that the company will use an age prediction system it is building to ensure that ChatGPT’s erotic features are only available to adult users.
As Altman wrote in a previous blog post, if OpenAI’s age prediction system incorrectly marks an adult as a minor, ChatGPT users may have to upload a photo of their government-issued ID to ChatGPT to correct it. Altman wrote that while this is an invasion of privacy, the company believes it is a “worthy trade-off.”
It’s unclear whether OpenAI will expand erotica into AI audio, image, and video generation tools.
Altman claims that OpenAI is making ChatGPT more friendly and erotic due to the company’s “treat adult users like adults” principle. Last year, OpenAI moved to a more lenient content moderation strategy for ChatGPT, making the chatbot more forgiving and offering fewer rejections. In February, OpenAI promised to express more political viewpoints in ChatGPT, and in March it updated ChatGPT to include AI-generated images of hate symbols.
These policies appear to be an attempt to make ChatGPT responses more popular with a variety of users. However, vulnerable ChatGPT users may benefit from safeguards that limit what chatbots can engage with. As OpenAI races towards 1 billion weekly active users, the tension between growth and protecting vulnerable users could become even more intense.
Updated 10/14/25 4:40pm PT: This story has been updated with comment from OpenAI.
