OpenAI is axing its NSFW adult feature ChatGPT that users begged for.
There's rarely a dull moment in the AI world right now. OpenAI CEO Sam Altman has been announcing his vision of AI being sold to consumers like a utility, while the company recently pulled the plug on its major video-creation project, taking its billion-dollar Disney partnership down with it in the process.
Although not everyone was upset about its discontinuation, as AI critics were quick to call the death of Sora a 'win' for humanity.
Meanwhile, the recent introduction of ads into ChatGPT's responses has frustrated users and drawn mockery from competitors like Anthropic, which has been happy to capitalise on the growing discontent.
Advert
Despite its setbacks, OpenAI isn't in decline; the AI giant recently secured an additional $10 billion in funding, nudging its valuation ever closer to the $1 trillion mark.
But that might be about to change as the company is axing its popular 'adult mode' chatbot that allows 'verified adult users' to engage in text-based erotica, according to the Financial Times.
OpenAI has reportedly confirmed that the feature is on hold indefinitely, citing a need for more time to assess the long-term implications of hosting that kind of content.
Given the broader conversation currently happening around AI and mental health, the concerns seem valid.
There have been growing fears about what's been described as AI psychosis, which is where some users have found themselves drawn into deeply unhealthy or delusional behaviours after intimate engagement with AI chatbots, with some cases ending in tragedy.
Earlier this month, the Wall Street Journal reported that company advisors had been raising serious red flags about the risks of allowing OpenAI's increasingly dependent user base to engage in intimate interactions with the technology.
Speaking to FT, one former senior employee said: “AI shouldn’t replace your friends or your family; you should have human connections.”
In a statement on 9 March, the company wrote that it was pushing back the launch of adult mode in order to concentrate on 'higher priority' developments including 'gains in intelligence, personality improvements, personalization, and making the experience more proactive.'
The statement added: “We still believe in the principle of treating adults like adults, but getting the experience right will take more time.”
At the same time, the WSJ revealed that OpenAI's age verification system had an error rate of over ten percent - a major flaw that could have left the door wide open for millions of underage users to access explicit content.