Adult Mode ‘ON’: OpenAI Lifts the Ban Under Competitive and Profit Pressures

OpenAI and Grok’s Expansion into Adult Content
Paid Model Aimed at Maximizing Revenue
Altman: “We Are Not the Moral Police”

Following xAI’s Grok, OpenAI’s ChatGPT has also begun allowing sexually explicit content for adults, marking a striking shift among major U.S. artificial intelligence (AI) platforms. While preventing the misuse of AI chatbots by children, adolescents, and those with mental health issues remains crucial, OpenAI argues that adult users should have the freedom to interact with AI in the ways they choose. The move reflects a strategic decision by startups seeking rapid monetization despite ethical controversy—and highlights how AI is evolving to turn human desire itself into a new market commodity.

ChatGPT to Permit Adult Content Starting in December

According to the information and communications technology (ICT) industry on October 20, OpenAI plans to begin offering adult-oriented content, including sexual conversations, to age-verified users on ChatGPT starting this December. Sam Altman, CEO of OpenAI, stated on October 17 (local time), “Under the principle of ‘treat adults like adults,’ we will allow much more—including erotic content—for users whose age has been verified.”

Altman noted, “We made ChatGPT quite restrictive to be cautious about mental health issues,” but added, “We realized that such restrictions made it less useful or less enjoyable for many users who don’t face those issues.” He continued, “We’ve now significantly mitigated the severe mental health risks and developed new tools, which means we can safely relax those restrictions in most cases.” Although specific content thresholds have not been disclosed, industry observers say the move is aimed squarely at rival xAI’s Grok, drawing attention to the level of explicitness Grok currently permits.

Launched in 2023 as an AI “that answers raw and unfiltered,” Grok expanded into sexual content this July. Its representative feature is Ani, a paid AI companion characterized as an anime-style female persona who engages in explicit, erotic conversations with users upon age verification, simulating romantic interaction. When generating images or videos, Grok—unlike services such as Sora 2—permits more explicit depictions for adults. Other platforms like Character.AI also allow users to engage in romantic roleplay and sexual dialogues with virtual characters.

Grok Expands User Base Through “Spicy Mode”

OpenAI’s introduction of adult mode is seen as following the broader trend among generative AI platforms like Grok, which have expanded their user bases through sexually explicit text, image, and video content. Before launching Ani, Grok had already rolled out a separate “Spicy Mode,” allowing users to generate adult content through text or image prompts.

xAI offers “Spicy Mode” to subscribers of its $30-per-month SuperGrok plan, which also includes AI image and video generation features. According to official documentation, images exceeding certain nudity thresholds are automatically blurred; however, in practice, many generated outputs feature translucent lace, lingerie, and near-nude depictions. “Spicy Mode” is not enabled by default—it is an optional NSFW (Not Safe for Work) content generation feature. Elon Musk has personally promoted the mode, sharing images generated with it on his X account.

Demand for AI that simulates romantic or sexual interaction clearly exists. A Harvard Business Review survey of 6,000 regular users in April found that “companionship therapy” was the most common use case. Research by investment firm Ark Invest shows that adult-oriented AI platforms’ market share surged from 1.5% in the previous year to 14.5%, encroaching on the space previously dominated by adult-content platform OnlyFans. The growing popularity of companion chatbots such as Character.AI and Replika further confirms this trend. Jessica Lee, senior fellow at Georgetown University’s CSET, observed, “OpenAI appears to have recognized this strong demand signal,” noting that erotic and adult content can also represent “a form of emotional immersion” in the human–model relationship context.

Furthermore, xAI’s earlier foray into adult content has already sent ripples through the broader market ecosystem. In particular, apps that had monetized through companion or interactive sexual chatbot modes for adult users have been significantly impacted. Experts view this as a signal that existing business models are being reassessed in the wake of large-platform entry into the segment.

Profitability vs. Ethics

Analysts believe the trajectory of this trend will hinge on the tension between profitability and ethical legitimacy. Technologically, AI-driven adult interaction is already feasible; the real challenge for platforms lies in social and regulatory backlash. Child protection organizations and industry officials warn that minors could bypass verification and gain exposure to adult material. Haley McNamara, vice president of the National Center on Sexual Exploitation, said in a statement, “Sexually objectified AI chatbots are inherently dangerous and could harm mental health by creating a false sense of intimacy.”

Entrepreneur Mark Cuban also criticized OpenAI’s plan, warning it “could face major blowback” and that parents would not trust OpenAI’s age filters. The U.S. Federal Trade Commission (FTC) has already launched an investigation into chatbot interactions with minors, while several states are considering stricter regulations on digital companions and sexual AI content. Jenny Kim, an attorney at Boies Schiller Flexner, told the BBC that “OpenAI is essentially using people as blind test subjects,” pressing the company on its safeguards against underage access.

In response, OpenAI said it is developing an “age prediction system” to determine whether a user is 18 or older. If age verification is uncertain, the default experience will be set to “under 18,” while adults will need to authenticate to unlock mature features. Altman responded on X, saying, “The announcement got more overhyped on the adult-content angle than intended.” He emphasized that the shift is “simply an expansion of freedom for adults, not a rollback of mental health guardrails or safety measures.” He added, “We are not the elected moral police of the world,” stressing that while safety must take precedence over privacy and freedom for minors, adults should have greater autonomy.

This shift, moreover, is not unprecedented. In February, OpenAI quietly revised its “Model Spec,” easing restrictions on violent and sexual content as part of a departure from what it termed “AI paternalism.” The update allowed the generation of sensitive textual material within appropriate contexts. The company says the upcoming policy change continues that direction—implementing the latest model specifications while maintaining bans on harmful applications such as deepfakes. Ultimately, the key question is how OpenAI will maintain equilibrium as it selectively opens its adult features.