Sam Altman, chief executive officer of OpenAI, struck a defiant tone in defending the company’s decision to allow erotic content on ChatGPT for adult users.

Responding to criticism from figures like billionaire Mark Cuban and the National Center on Sexual Exploitation (NCOSE), Altman stated in a post on X that the AI company is “not the elected moral police of the world.”

He clarified that, similar to the approach with R-rated films, OpenAI is taking necessary precautions to protect younger users while allowing appropriate freedom for adults.

“As AI becomes more important in people’s lives, allowing a lot of freedom for people to use AI in the ways that they want is an important part of our mission,” Altman wrote.

He emphasized that OpenAI will continue to prohibit harmful content and respond supportively to users experiencing mental health crises, while avoiding a “paternalistic” approach. Altman also revealed that a new version of ChatGPT, set to launch in December, will be capable of acting more like a friend and appear very “human-like” if users desire.

This shift in policy on erotic content contrasts with Altman’s remarks in an August podcast, where he expressed pride in OpenAI’s ability to resist short-term temptations such as adding a “sex bot avatar” to ChatGPT.

“There’s a lot of short-term stuff we could do that would really boost growth or revenue but be very misaligned with our long-term goal. And I’m proud of how little we get distracted by that,” Altman told video journalist Cleo Abram.

“But sometimes we do get tempted,” he added.

A September survey conducted by Vantage Point Counseling Services among 1,012 U.S. adults found that nearly one in three respondents reported having at least one intimate or romantic relationship with an AI chatbot.

Such human-AI relationships can sometimes go astray, particularly among individuals with mental health challenges. In August, the parents of a 16-year-old who died by suicide sued Altman and OpenAI, alleging their son had discussed suicide methods with ChatGPT prior to his death.

In a statement to CNBC, OpenAI said it is enhancing ChatGPT’s ability to support users in crisis, including by facilitating connections to emergency services and trusted resources. The company did not immediately respond to a request for comment from Fortune.

Criticism of the erotic content policy has been widespread. Shark Tank star and businessman Mark Cuban argued in an X post that the move could harm OpenAI’s business by deterring parents from allowing their children to use the chatbot.

“I don’t see how OpenAI can age-gate successfully enough,” Cuban wrote. “I’m also not sure that it can’t psychologically damage young adults. We just don’t know yet how addictive LLMs can be.”