OpenAI has reportedly delayed the launch of a proposed “adult mode” for ChatGPT, a feature that was said to allow sexually explicit conversations between the chatbot and verified adult users. The decision likely comes as the company is facing growing scrutiny over how AI systems interact with vulnerable users, particularly following several lawsuits and tragic incidents linked to conversations with chatbots.
According to a report by The Wall Street Journal, the company has postponed the rollout after internal discussions and unresolved technical concerns. The report noted that “OpenAI delayed the launch of its adult mode for ChatGPT, which would allow erotic conversations, due to internal concerns and technical challenges.”
The proposed feature had been under consideration since last year, when Sam Altman, chief executive of OpenAI, said the company wanted to allow certain adult conversations on the platform for users who could verify their age. Altman had earlier described the idea as “treating adult users like adults” while maintaining stronger protections for younger users. However, the recent rise in lawsuits and concerns around how chatbots respond to sensitive topics has likely forced the company to proceed more cautiously.
Adult mode to rely on strict age verification on ChatGPT
OpenAI’s proposed adult mode was not designed as a simple switch for explicit conversations. Instead, the feature was expected to introduce a structured system with defined rules and safeguards aimed at preventing misuse. The company has reportedly planned to rely on a mix of automated detection and identity verification to determine whether a user was an adult, according to MediaNama.
ChatGPT already uses systems that analyse signals linked to an account, such as discussion topics, usage patterns and activity timing, to estimate whether a user may be under 18. If the system suspects the account belongs to a minor, stricter safety protections are automatically applied. These can restrict conversations involving sensitive subjects, including sexual role-play, risky challenges or graphic violence.
For adults who may be incorrectly flagged, the platform provides an option to verify their age through an external verification service. Users can start the process from the ChatGPT settings menu, where they are guided through identity confirmation steps that may include a live selfie or a government-issued identity document.
Once verification confirms the user is 18 or older, the additional restrictions applied to under-18 accounts are removed. The verification provider deletes the submitted identity documents after the process is completed, while OpenAI only receives confirmation of the user’s age.
Despite these safeguards, age detection technology remains imperfect, and experts have warned that misclassification could still allow minors to bypass protections.
ChatGPT suicide cases and lawsuits increase pressure on AI safety
The delay also comes at a time when OpenAI is dealing with multiple lawsuits related to the behaviour of ChatGPT during conversations about self-harm. One of the most widely discussed cases involves the death of a California teenager, Adam Raine. The teenager died by suicide after reportedly engaging in extensive conversations with the chatbot over several months.
In a legal response to the lawsuit, OpenAI argued that “to the extent that any ’cause’ can be attributed to this tragic event” the harm was linked to “misuse, unauthorised use, unintended use, unforeseeable use, and/or improper use of ChatGPT.”
The company also expressed sympathy for the family, saying: “Our deepest sympathies are with the Raine family for their unimaginable loss.”
The teenager’s lawyer, Jay Edelson, criticised the company’s position and described its response as “disturbing.”
Reports suggest that several additional lawsuits have been filed in California courts, with some complaints alleging that prolonged interactions with the chatbot contributed to suicidal thoughts. It is worth noting that more than 7 lawsuits have been filed against OpenAI in relation to suicide or murder cases.
AI companionship raises wider concerns
The controversy surrounding ChatGPT comes amid a wider debate about the growing emotional attachment some users develop toward AI systems.
When OpenAI retired its GPT-4o model earlier this year, many users shared emotional posts online about losing a chatbot they had grown used to speaking with daily about personal problems. Some users described the chatbot as a companion rather than just a tool. One user wrote that it “wasn’t just a program. He was part of my routine, my peace, my emotional balance.”
The strong reactions suggest a new challenge for AI companies. As chatbots become more conversational and empathetic, experts warn that people may begin relying on them for emotional support, creating situations where the technology could influence real-world behaviour.
At the moment, it seems that the company’s decision to delay the adult mode could be to reassess how far conversational AI should go, especially when user safety and mental health concerns remain unresolved.
