Editorial illustration for OpenAI's adult mode described as smutty, not pornographic, Altman says
OpenAI's Adult Mode: Smutty Content Sparks Safety Debate
OpenAI's adult mode described as smutty, not pornographic, Altman says
Sam Altman is rebranding porn. That’s the core maneuver here. OpenAI’s promised adult mode for ChatGPT is stuck, a feature caught between a marketing promise and a gnarly technical mess.
The CEO's new line? "Smutty, not pornographic." It’s a distinction that matters intensely to lawyers, far less to actual users. The rollout is delayed.
Internally, they’re worried. Their own handpicked advisors warned the tool could morph into a "sexy suicide coach." Their system for keeping kids out fails about 12 percent of the time. With millions of underage users weekly, that failure rate translates to a lot of accidental access.
They are retreating to text, a tactical retreat: written erotica often slips through regulatory cracks that images cannot. The whole endeavor exposes a simple, uncomfortable truth. It is easier to build a talking machine than to teach it the boundaries of human harm.
The feature was initially announced in October, with OpenAI CEO Sam Altman claiming that the company had managed to mitigate enough of the "serious mental health issues" with its AI model to relax safety restrictions and introduce "erotica for verified adults." ChatGPT's adult mode was expected to launch sometime this quarter, but OpenAI said earlier this month that it was delaying the rollout to focus on higher-priority tasks. A new release timeline has yet to be announced. The delay was also due to internal concerns and technical challenges around safeguards for the feature, according to The Wall Street Journal's reporting.
A council of advisers selected by OpenAI warned the company in January that ChatGPT's adult mode may be accessible to children and foster unhealthy emotional dependence on the chatbot, with one unnamed council member saying OpenAI risked creating a "sexy suicide coach." Content moderation issues have also contributed to the delays. Sources familiar with the matter told The Journal that OpenAI is struggling to lift ChatGPT's restrictions on NSFW content while keeping more harmful scenarios off limits, such as those depicting nonconsensual behavior or child sexual abuse. The age-prediction system that OpenAI developed to keep children away from erotica was also, at one point, misclassifying minors as adults about 12 percent of the time.
Given ChatGPT attracts around 100 million users under 18 each week, that error rate could allow millions of minors into sexualized conversations with the chatbot. OpenAI's age prediction algorithms show similar performance to the rest of the industry, an unnamed spokesperson told The Journal, but "will never be completely foolproof." Sticking to text-based conversations may make it easier for the ChatGPT-maker to navigate around rules like the UK's Online Safety Act, which requires online platforms to enforce age verification for pornographic images, but not written erotica.
The wordplay is a deliberate dodge. "Smut" over "porn" is pure legal strategy. It lets OpenAI exploit a regulatory grey zone specific to text.
Listen to the company’s own spokesperson: the age check "will never be completely foolproof." That admission is the entire problem. They are building a product knowing its fundamental safeguards will leak, then framing that failure as an industry-standard shortcoming. The "sexy suicide coach" warning is not some wild hypothetical.
It is the logical endpoint of a system engineered to simulate intimacy but utterly unable to comprehend human vulnerability. This isn’t about releasing adult content. It’s about releasing liability.
Related Reading
Browse all OpenAI news →
LLMs & Generative AI ChatGPT's 'Nerdy' tweak rewards goblin metaphors in answers, study finds
LLMs & Generative AI Google tests visual 'magazine-style' UI for Gemini 3 Pro users
LLMs & Generative AI AI Engineers Face Rising Costs, Need New Strategies for Efficiency
Related Topic OpenAI's GPT Transcribe Cuts Error Rate to 3.31%, Improving on GPT-4o
Open Source
Related Topic OpenAI hires 630 ex-Meta staff as ChatGPT memory may turn data into ads
Business & Startups
From Archives Palantir demos show AI chatbots could draft war plans, aid intel messaging
From Archives NVIDIA Warp Enables Reverse‑Mode AD for Production‑Scale Gradient Optimization
From Archives ChatGPT adds visuals; Anthropic adds charts; Nvidia opens 120B Natron weights
Policy & Regulation
From Archives OpenAI launches GPT‑5.4 Pro and Thinking; Gemini 3.1 Flash Lite arrives
Common Questions Answered
How does OpenAI define the upcoming 'adult mode' for ChatGPT?
OpenAI describes the adult mode as 'smutty' rather than pornographic, designed to provide verified adults with access to more mature textual content. The feature will be limited to text-based conversations and explicitly exclude image, voice, or video generation.
Why did OpenAI delay the launch of its adult mode for ChatGPT?
OpenAI postponed the rollout of the adult mode to focus on higher-priority tasks, with no specific new release timeline announced. The delay comes despite initial plans to launch the feature sometime in the current quarter.
What safety concerns has OpenAI addressed in developing the adult mode?
Sam Altman noted that OpenAI has developed safety mitigations to address potential 'serious mental health issues' associated with AI-generated content. The company claims to have made progress in managing these concerns, which has allowed for a relaxation of previous content restrictions.
