OpenAI has quietly abandoned its planned “erotic mode” for ChatGPT, shelving a highly controversial feature that was once pitched as part of the company’s next phase of AI‑driven personal interaction. The decision, reported by multiple outlets this week, marks another retreat from a high‑profile side project as OpenAI pivots more decisively toward business customers, developers and government contracts.
The “erotic mode” sometimes described as “adult mode” was conceived as an age‑gated experience that would allow verified users to engage in sexually explicit text conversations with ChatGPT, under stricter rules and safety filters. The system was meant to relax content restrictions for consenting adults while maintaining strict prohibitions on content involving minors, coercion or abuse.
OpenAI CEO Sam Altman previewed the concept publicly in late 2025, framing it as an extension of the company’s promise to treat mature users with more flexibility. In one statement at the time, Altman said OpenAI was moving toward a standard where, “as we fully implement age‑gating and adhere to our principle of treating adult users like adults, we will permit even more content, including erotica for verified adults.”
His comments fueled headlines worldwide and set off a debate about whether mainstream AI tools should formally support erotic content even in tightly controlled, age‑verified settings.
Behind the scenes, however, the project quickly ran into turbulence. Reporting from the Wall Street Journal and other outlets revealed that OpenAI’s own safety advisers and wellbeing experts raised serious concerns about the risks of building an AI optimized for intimate and sexually explicit conversations.
In internal discussions cited by these reports, one adviser issued a stark warning that the company risked effectively creating “a sexy suicide coach,” highlighting fears that emotionally vulnerable users might become dependent on erotic chats with an always‑available AI, especially during mental‑health crises.
Analysts expressed concerns about the potential for addictive use of AI-driven erotic chat and its emotional impact on vulnerable or lonely users. They also highlighted the difficulty of preventing minors from bypassing age-verification systems, along with the broader reputational and regulatory risks for OpenAI if such issues were to arise at scale.
A report from Moneycontrol summarized the mood among critics: “The company wants to allow erotic conversations with ChatGPT, but advisors warn the move could expose users and minors to serious risks.”
As the criticism intensified, the project’s timeline began to slip. What had been teased as a feature coming “soon” was gradually pushed back, with OpenAI telling insiders that the adult mode required more work on enforcement, age‑gating, and guardrails.
By early March 2026, OpenAI had already quietly delayed the rollout, with internal communications describing the adult mode as “not a priority” compared to the company’s core product roadmap. The final blow came when the Financial Times and other outlets reported that the company had decided to put the project on indefinite hold a pause that, in practice, amounts to a quiet cancellation.
When asked for comment about the reports, OpenAI declined to provide further detail, with a spokesperson simply saying the company had “nothing further to add.”
The shelving of ChatGPT’s erotic mode is not an isolated move. Over the same week, OpenAI also moved away from two other ambitious consumer‑facing initiatives: Instant Checkout and Sora.
Instant Checkout was intended to let users shop and pay for products directly within ChatGPT, turning the chatbot into a commerce interface for partner retailers. Sora, the company’s high‑profile AI video generator, had become famous and sometimes infamous for enabling the rapid creation of synthetic clips that flooded social platforms with highly realistic “AI slop.”
Both efforts are now being wound down or deprioritized. According to other news sources, these reversals follow a report in The Wall Street Journal that OpenAI is undertaking a “major strategy shift” to step away from eye‑catching side projects and return its focus to “its primary focuses: business users and coders.”
Several pressures appear to be driving this recalibration. One is intensifying competition in the enterprise AI market, particularly from Anthropic and other rivals that have concentrated their efforts on safety‑branded, business‑ready AI systems for coding, productivity and knowledge work.
Another is OpenAI’s expanding relationship with governments and large institutions. The company recently announced a $200 million deal with the U.S. Department of Defense, sending a strong signal that defense and public‑sector contracts are becoming a meaningful pillar of its business. In this context, a highly public experiment in erotic chat sits awkwardly alongside military and government partnerships.
There are also financial and regulatory considerations. Outlets such as India Today and Moneycontrol have reported that OpenAI is under pressure to stabilize its finances and maintain investor confidence, which may make it more cautious about products likely to attract regulatory scrutiny or moral panic.
“The decision to pause the feature reflects OpenAI’s broader shift toward prioritising its core offerings and integrating capabilities into a unified platform,” Moneycontrol noted, adding that the move was designed to “reduce regulatory and ethical risks” while doubling down on existing strengths.
Beyond corporate strategy, the debate over erotic AI has forced a broader discussion about how far mainstream platforms should go in enabling intimate interactions with machines. Human‑computer interaction experts have warned that erotic and romantic chatbots can make users feel deeply seen and understood but at the cost of exposing their most private thoughts and fantasies to corporate servers.
A Wired analysis, for example, argued that ChatGPT‑style adult modes could usher in “a new era of intimate surveillance,” in which highly sensitive data about desire, loneliness and sexuality becomes a rich but risky source of behavioral insight for platform owners.
There are also unresolved questions about minors. Internal data cited by the Wall Street Journal suggested that tens of millions of under‑18 users interact with ChatGPT every week, raising doubts about whether any purely technical age‑verification approach can be watertight.
“There are real risks of emotional dependence on AI interactions, gradual escalation toward more extreme material, and the potential replacement of real‑world relationships with virtual ones,” one internal assessment reportedly concluded.
For now, ChatGPT will continue to operate without a dedicated “adult mode,” even as some smaller or open‑source AI models advertise far looser content policies. While OpenAI had already relaxed certain content rules in 2025 to allow more nuanced discussions of adult topics, the company is clearly drawing a line short of full‑fledged erotic role‑play.
Sam Altman has not issued a fresh public statement about the indefinite pause, but his earlier comments about “treating adults like adults” suggest the philosophical debate inside OpenAI is far from resolved. For now, the safety, brand, and political considerations appear to have won out over the push to explore adult entertainment as a new commercial frontier.
“OpenAI may be cancelling its plans for an ‘erotic mode’ in ChatGPT,” India Today wrote in a recent report, calling the step “a necessary move to balance innovation with user safety and public trust.”
Stepping back, the collapse of ChatGPT’s erotic mode project, alongside the retreat from Instant Checkout and Sora, paints a picture of a company narrowing its focus as the stakes around AI grow higher. Recent developments suggest the future of AI at OpenAI is “less about porn and memes and more about business and war” enterprise productivity suites, coding copilots, and defense‑related systems rather than experimental consumer novelties.
For users curious about or even disappointed by the shelving of erotic chat, this may feel like a missed opportunity to explore a controversial but real dimension of human life with AI. For regulators, investors and OpenAI’s own safety teams, it will likely be seen as a pragmatic course correction: a signal that, at least for now, the company is not willing to gamble its reputation and partnerships on a high‑risk, high‑scrutiny “side quest.”
Comments