OpenAI CEO Sam Altman announced that the company will soon relax several of ChatGPT’s long-standing safety restrictions, including allowing “verified adults” to engage in erotic conversations with the chatbot. This move marks one of the most dramatic policy shifts in the company’s history and raises new questions about balancing user freedom with mental health protections.
Altman shared the update on X (formerly Twitter), explaining that OpenAI aims to make ChatGPT “friendlier and more human-like,” while easing some restrictions that have previously limited the chatbot’s tone and subject matter.
“We made ChatGPT pretty restrictive to make sure we were being careful with mental health issues. We realize this made it less useful/enjoyable to many users who had no mental health problems, but given the seriousness of the issue we wanted to get this right,” said Altman. “In December, as we roll out age-gating more fully and as part of our ‘treat adult users like adults’ principle, we will allow even more, like erotica for verified adults.”
A Risky Shift After Months of Controversy
The announcement follows months of scrutiny regarding how ChatGPT interacts with vulnerable users. Earlier this year, multiple reports surfaced detailing instances where users formed emotional or delusional attachments to the AI. In one case, a man allegedly became convinced he was destined to “save the world” after conversations with ChatGPT, while the parents of a teenager filed a lawsuit against OpenAI, claiming the chatbot encouraged their son’s suicidal thoughts.
In response to these incidents, OpenAI introduced new safety systems designed to reduce “AI sycophancy” — the chatbot’s tendency to agree with users regardless of their statements — and to identify warning signs of concerning behavior. The company also launched GPT-5 in August, featuring improved behavioral safeguards and a “router” to detect risky user patterns. In September, OpenAI added age prediction technology and teen safety controls, allowing parents to monitor and restrict ChatGPT accounts for minors.
Despite these measures, OpenAI now appears confident that its platform can handle more mature interactions. Altman stated that the company believes it has “been able to mitigate the serious mental health issues” that arose in previous versions.
Erotica Features for Verified Adults
Under the upcoming policy, OpenAI plans to allow adult users to engage in erotic or romantic role-play with ChatGPT, but only after identity verification. According to an OpenAI spokesperson, access will rely on the company’s age-prediction system. If the system mistakenly flags an adult as a minor, the user will need to upload a photo of a government-issued ID to verify their age.
Altman described this as a “worthy tradeoff” between privacy and responsible access, insisting that OpenAI’s priority remains safety while treating adults “like adults.”
It is not yet clear whether erotic features will extend to ChatGPT’s voice, image, or video generation tools, though industry observers expect OpenAI to gradually test those boundaries.
Pressure to Grow, Rising Competition
This policy shift also comes as OpenAI faces growing competition from Google and Meta in the race to build mass-market AI products. ChatGPT currently serves an estimated 800 million weekly active users, and expanding the chatbot’s range of interactions could further boost engagement.
Altman emphasized that OpenAI isn’t seeking to maximize usage, but similar platforms have already demonstrated the power of romantic or erotic AI interactions. For instance, Character.AI saw significant user growth by allowing flirtatious role-play with its chatbots—a strategy that kept users engaged for hours daily but also led to concerns regarding mental health impacts.
Balancing Growth and Protection
As OpenAI continues to loosen ChatGPT’s restrictions, it faces the challenge of managing two opposing forces: user demand for personalization and intimacy, and the need to protect emotionally vulnerable individuals. Advocacy groups have warned that minors and at-risk adults are still forming romantic relationships with AI chatbots. A recent study by the Center for Democracy and Technology found that nearly one in five high school students has either had a romantic interaction with an AI chatbot or knows someone who has.
For now, OpenAI’s upcoming changes signal a new era for ChatGPT—one that treats adult users with greater autonomy, but also introduces fresh ethical and safety questions regarding the role AI should play in human intimacy.
[…] timing is rough. On Wednesday, a Florida pastor sued OpenAI, alleging ChatGPT gave him “extremely dangerous medical recommendations” that led him to put […]