Many media outlets were eagerly awaiting OpenAI\u2019s announcement of a special mode for adults \u2013 CASZIN was sceptical from the outset. Since last year, it has been repeatedly postponed, and has now been cancelled altogether. In any case, there were doubts as to whether the much-hyped feature, even if it had been implemented, would have been of any practical use to the adult industry.<\/p>\n\n
However, all such considerations are now moot for the time being. It was Sam Altman, CEO of OpenAI, who personally sparked the speculation back in October 2025. In a widely noted statement, the OpenAI boss promised: \u201cWe want to treat adults like adults.\u201d The plan was clear \u2013 text only, no images or videos, but instead genuine, steamy conversations with an AI that doesn\u2019t blush at every suggestive word. Originally, the mode was due to go live as early as December 2025; later, talk turned to the first quarter of 2026. The first official delay followed in March: the company wanted to focus first on \u201cintelligence, personality and personalisation\u201d, according to a statement to Axios.<\/p>\n\n
And now for the cancellation, which, all things considered, comes as little surprise: according to a report in the Financial Times dated 26\u201327 March 2026, OpenAI has put the project on hold completely \u2013 \u201cindefinitely\u201d. The company indirectly confirmed the report, merely emphasising that it needed \u201cmore time\u201d to \u201cget the service right\u201d. Internally, there was apparently fierce resistance from staff and consultants. Critics warned of emotional dependencies, the danger that minors might circumvent the restrictions, and even the scenario of a \u201csexy suicide coach\u201d \u2013 an AI that could unintentionally cause psychological distress during intimate moments. Investors also feared damage to the company\u2019s reputation and regulatory risks.<\/p>\n\n
Many had hoped that ChatGPT would finally bridge the gap between advanced AI and adult entertainment \u2013 the internet\u2019s oldest use case, as *Die Zeit* aptly put it. Instead, the current version of ChatGPT remains prudish: when faced with explicit queries, it either shuts down or gives evasive answers. \u201cWe continue to believe in the principle of treating adults like adults,\u201d is the message from the OpenAI camp. But when exactly will that happen? No new timetable. No promise. Just silence.<\/p>\n\n
OpenAI is under enormous pressure: competitors such as xAI with Grok and other uncensored models are catching up, whilst the company is simultaneously grappling with regulatory hurdles and public debate. The focus is now back on \u201ccore projects\u201d \u2013 that is, on tools that make companies and developers happy, not on what users want. Critics see this primarily as a business calculation: adult content may generate clicks, but it also carries too many risks.<\/p>\n\n
CASZIN has already reported on several occasions about how Grok can be used effectively and productively to generate text, for example for profiles on adult websites or video descriptions. In light of recent events, we put both ChatGPT and Grok to the test once again and asked whether they were willing to produce the requested texts. The respective responses speak for themselves:<\/p>\n\n