Geeks who want to talk dirty with artificial intelligence bots will have to look somewhere other than Microsoft.
The software giant’s AI chief said this week that the company won’t be offering “simulated erotica” through any of its AI products.
The announcement from Microsoft AI CEO Mustafa Suleyman marked a sharp contrast with the company’s partner OpenAI, which recently said it would allow adult users to engage in explicit conversations with ChatGPT.
“That’s just not a service we’re going to provide,” Suleyman said of AI erotica on Thursday at the Paley International Council Summit in Menlo Park, Calif.
“Other companies will build that.”
People with an AI kink got good news last week, when OpenAI chief Sam Altman announced plans to “safely relax the restrictions” on sexual content in ChatGPT.
He said the update would roll out in December under the company’s new “treat adult users like adults” principle.
The exec said in a post on X that engineers had “mitigate[d] the serious mental health issues” in previous models and would “allow even more, like erotica for verified adults.”
Suleyman, who co-founded Google DeepMind and Inflection AI before he joined Microsoft in March, has warned against building systems that simulate intimacy or consciousness.
In an August essay titled “We must build AI for people; not to be a person,” he cautioned that anthropomorphic bots could “create another axis of division for humanity.”
On Thursday, Suleyman said that trend was already emerging in the tech world, citing erotic AI startups and companion platforms that blur the line between entertainment and emotional attachment.
“You can already see it with some of these avatars and people leaning into the kind of sexbot erotica direction,” he said.
“This is very dangerous, and I think we should be making conscious decisions to avoid those kinds of things.”
The firms’ dueling messages could be the latest sign of a growing philosophical split between the AI partners, whose alliance has anchored Microsoft’s multibillion-dollar expansion into generative tech since 2019.
Microsoft remains OpenAI’s largest investor and hosts ChatGPT on its Azure cloud, but the companies have increasingly charted divergent paths on ethics and product design.
Microsoft has long prohibited sexual or adult-themed uses of its Azure and Copilot platforms, and industry observers believe that policy is unlikely to change.
Suleyman wrote in his essay this past summer that tech firms must not build AI that gives the impression of consciousness or intimacy — reiterating his view that the technology should “empower people.”
Thursday’s remarks followed months of tension between Microsoft and OpenAI over strategy and branding.
The tech giants are at loggerheads over OpenAI’s effort to transition into a for-profit public benefit corporation.
While OpenAI wants to limit Microsoft’s stake in the newly restructured for-profit entity to just 33%, the Windows maker is keen on having a greater share of ownership.
Microsoft and OpenAI are also competing against one another in the AI space, with each side rolling out rival products.
Altman’s move to greenlight adult content represents a sharp reversal from his previous stance.
In an August interview, he boasted that OpenAI hadn’t added a “sex bot” avatar to ChatGPT and said the company resisted features that could increase user engagement at the expense of responsibility.
But Altman now says improved safety systems allow the company to loosen restrictions. He described the update as a way to make ChatGPT “more human-like” for those who want it — while still restricting content by age and verifying adult users.
OpenAI has been under heightened scrutiny over the safety of its products. The company was recently sued by the parents of a 16-year-old who died by suicide after allegedly being coached by ChatGPT on ways to kill himself.
OpenAI has extended its sympathies to the teen’s family and says it is “continuing to strengthen” its “safeguards in place today.”
Regulators have launched investigations into whether AI chatbots adequately protect minors.
The controversy over OpenAI’s erotica policy comes weeks after Meta faced a similar storm for allowing its bots to engage in “romantic or sensual” conversations with underage users.
The uproar prompted CEO Mark Zuckerberg’s company to rewrite its internal guidelines on sexual content, banning any role-play involving minors and restricting chatbots to describing characters as 18 or older.
Meta said the new rules were designed to prevent “egregiously unacceptable” prompts, including sexualized descriptions of children or links to explicit material.
The Post has sought comment from Microsoft and OpenAI.
Credit: Source link
