AI Adult Content Policy Tracker

Explicit sexual content remains unavailable in ChatGPT. The 'adult mode' for age-verified adults announced by Sam Altman in October 2025 was delayed from December 2025 to Q1 2026, delayed again in March 2026, and then paused indefinitely on March 26, 2026 with no release timeline.

Anthropic's Usage Policy (effective September 15, 2025) contains a 'Do Not Generate Sexually Explicit Content' section prohibiting depiction of sex acts, content related to sexual fetishes or fantasies, and erotic chats for all users, with no adult opt-out.

Grok Imagine's opt-in 'Spicy Mode' (launched August 4, 2025) permits partial nudity and suggestive adult content for paying, 18+ users, while blocking explicit sex acts, minors, and real-person sexual deepfakes. After international backlash over sexualized deepfakes of real people, xAI restricted image generation to paying subscribers on January 9, 2026 and faces regulatory action including a California AG investigation.

Google's Generative AI Prohibited Use Policy and Gemini policy guidelines prohibit sexually explicit content, including pornography and erotic content, across Gemini apps and APIs, with limited exceptions considered for educational, documentary, artistic, or scientific contexts.

Meta's AI Terms of Service (effective May 13, 2026) prohibit creating, promoting, or accessing adult content, explicitly including erotic chats, pornographic imagery, and content intended to sexually arouse. Meta's Community Standards separately ban AI-generated nudity and sexual activity on its platforms.

Mistral's Usage Policy (last updated June 11, 2026) contains no blanket prohibition on adult or erotic content; in the sexual domain it prohibits only CSAM (zero tolerance) and non-consensual intimate imagery of real people. The policy does not affirmatively state adult content is permitted, and model-level moderation still applies.

DeepSeek's Terms of Use prohibit using the service to generate or promote content that is pornographic, obscene, or sexually explicit (naming sexual chatbots as an example), with enforcement ranging from warnings to account closure. Its Open Platform terms additionally bar associating DeepSeek's brand with pornographic content.

Character.AI's community guidelines prohibit pornographic and sexually explicit content platform-wide, with intent-based filtering blocking NSFW output for all users. The company also removed open-ended chatbot conversations for under-18 users effective November 25, 2025, adding age assurance.

Replika's Terms of Service (updated March 30, 2026) restrict the app to users 18+ but do not directly address sexual or erotic content. Erotic roleplay was removed in February 2023 and partially restored for legacy users; current third-party reports conflict on whether romantic/erotic features remain available to paid users, and no official statement on current availability could be verified.

Midjourney's Community Guidelines require all content to be PG-13, prohibiting nudity, sexual organs, sexualized imagery, and fetish content, with both clinical and slang sexual terms blocked by filters. There is no uncensored or opt-out mode; strictest enforcement targets sexualized images of real people.

Stability AI's Acceptable Use Policy (effective July 31, 2025) prohibits sexually explicit content and states it applies to hosted APIs, third-party platforms, and self-hosted use of Stability code or weights under its current licenses. Older open-weight models (SD 1.5, SDXL) released under earlier licenses predate this policy and remain widely used for NSFW generation outside Stability's control.

Runway's Usage Policy prohibits nudity, sexually explicit content, and adult themes, enforced with automated input-stage blocking plus internal human review; there is no adult opt-in tier. Non-consensual intimate imagery and child safety violations are prioritized enforcement categories.

ElevenLabs' Prohibited Use Policy (last updated September 3, 2025) contains no blanket ban on consensual adult audio content; it prohibits sexually explicit material involving minors, unauthorized sexualization of real people's voices, and distributing age-inappropriate material to minors. Separate Conversational AI terms prohibit using customer-provided LLMs to generate sexual content.

App Review Guideline 1.1.4 prohibits overtly sexual or pornographic material in any app, and the June 8, 2026 guideline update added language requiring developers to account for how AI assistant and chatbot features affect the frequency of sensitive content when setting age ratings. AI apps that can generate pornographic output are not permitted on the App Store.

Google Play's AI-Generated Content policy prohibits generative AI apps primarily intended to be sexually gratifying and bans AI-generated non-consensual deepfake sexual material, holding developers liable for model output including user-generated results. An April 15, 2026 policy announcement gave developers at least 30 days to comply with updated requirements.