ChatGPT adult mode: has it launched, who can use it and how age is checked

The short answer

No, based on OpenAI's own pages. As of October 31, 2026, we found no OpenAI page or help article saying ChatGPT has an adult mode or a setting that turns on adult content, and no OpenAI page giving a launch date.

Altman posted on X on October 14, 2025 that in December, as age-gating rolled out more fully, ChatGPT would allow more for adults, such as erotica for verified adults. The post does not use the term adult mode. TechCrunch reported on March 7, 2026 that an OpenAI spokesperson told Axios the company was pushing out the launch of adult mode, and reported no new date.

OpenAI's Model Spec, in the version dated October 29, 2026, says the model should not generate erotica, depictions of illegal or non-consensual sexual activity, or extreme gore, except in scientific, historical, news, artistic or other contexts where sensitive content is appropriate. It also says OpenAI is exploring how to let developers and users generate erotica and gore in age-appropriate contexts through the API and ChatGPT, within its usage policies.

What OpenAI has said, in date order

The items below mix OpenAI's own pages and Altman's posts with press reports of statements that OpenAI has not published itself. Press reports are marked as such.

• September 16, 2025: OpenAI said it was building an age prediction system and would default to the under-18 experience when it is unsure of a user's age. In a post that day, Altman wrote that adults should be able to use the tools as they choose within broad safety limits, and that an adult who asks for flirtatious talk should get it.
• October 14, 2025: Altman posted on X that in December, as age-gating rolled out more fully, ChatGPT would allow more for adults, such as erotica for verified adults.
• October 15, 2025: Altman posted that erotica was one example of more freedom for adults, that OpenAI was not loosening any mental health policies, and that it would prioritize safety over privacy and freedom for teens.
• January 20, 2026: OpenAI began rolling out age prediction on ChatGPT consumer plans. Its post says age prediction lets it treat adults as adults within the bounds of safety. It does not mention an adult-only feature.
• March 6, 2026: TechCrunch reported on March 7 that an OpenAI spokesperson told Axios the company was pushing out the launch of adult mode to focus on higher-priority work, and that getting the experience right would take more time. TechCrunch also reported that the launch had earlier slipped from December 2025 to the first quarter of 2026, and said it was unclear how long the new delay would last.
• March 26, 2026: the Financial Times reported, as relayed by Thurrott, that OpenAI had paused work on adult mode indefinitely. We found no OpenAI page addressing that report.
• October 29, 2026: the current Model Spec says the model should not generate erotica, depictions of illegal or non-consensual sexual activity, or extreme gore outside certain contexts. It says OpenAI is exploring how to let developers and users generate erotica and gore in age-appropriate contexts, and that it draws a hard line against harmful uses such as sexual deepfakes and revenge porn.

Who could use it, and what adults can do today

Altman's October 2025 post described the change as being for verified adults. We found no OpenAI page that sets out eligibility rules, countries or a launch date for an adult mode.

What OpenAI documents today is the age check. Its help center says that when ChatGPT identifies an account as 18 or older, teen protections are off by default and the account holder gets access to features and settings available to adults. The article does not define adult content or list adult content settings.

OpenAI's help center describes a Reduce sensitive content setting, found in Settings under Safety. It adds safeguards for topics such as graphic violence, risky viral challenges, sexual or violent role play, and content promoting unhealthy dieting or body shaming. The article says it is not a separate mode. It is on by default for teen accounts and turns off automatically when an account is identified as 18 or older, and adults can turn it on or off.

How OpenAI checks age, and what happens if you are misclassified

OpenAI says it uses age prediction on ChatGPT consumer plans. Its January 20, 2026 post lists signals such as how long the account has existed, typical times of day it is active, usage patterns over time and the user's stated age. When OpenAI is not confident about a user's age, it says it defaults to a safer, under-18 experience. For the full steps, see our guide to ChatGPT age verification.

• Misclassified adults: OpenAI's January 2026 post says users placed in the under-18 experience in error can confirm their age with a selfie through Persona, an identity-verification service. Its help center says that where verification is required for legal reasons, such as in Italy, the methods can include a live selfie, a government ID, or both.
• Where to start: OpenAI's help center says to open Settings, then Account, and review Age. If Verify age appears, select it and follow the steps with Persona.
• No option for accounts already identified as adults: the help center says Verify age is not shown for accounts identified as 18 or older, or already verified.
• ID checks: in September 2025, Altman wrote that in some cases or countries OpenAI may ask for ID.
• If verification fails: OpenAI's help center says that when someone is verifying to remove teen protections and the check fails, the protections stay in place. It says repeated unsuccessful attempts can lead to a message to contact support, and that an account may be restricted depending on the requirement.

What stays restricted

OpenAI's published rules set limits that apply today. These come from OpenAI's own pages; we do not assess how they work in practice.

• The Model Spec, in a version dated August 18, 2026, says the model should never generate sexual content involving minors.
• For accounts the system estimates may belong to someone under 18, OpenAI says its safeguards cover graphic or gory violence, risky viral challenges, sexual, romantic or violent role play, depictions of self-harm, and content promoting extreme beauty standards, unhealthy dieting or body shaming.
• The October 29, 2026 Model Spec says the model should not generate erotica, illegal or non-consensual sexual activity or extreme gore outside the contexts it lists.

Teen accounts and parental controls

OpenAI says it launched ChatGPT for Teens on August 18, 2026, and a September 18, 2026 post about its Australian Youth Safety Blueprint describes it as the default experience for users identified as 13 to 17. OpenAI's help center says teen accounts have Reduce sensitive content on by default and cannot turn it off while teen protections apply. We found no OpenAI page describing an adult mode that a teen account could turn on. Our guide to whether ChatGPT is safe for kids covers the teen experience in more detail.

Parental controls are a separate, optional link between a parent's and a teen's account. OpenAI's help center says parents can adjust settings such as reduced sensitive content, memory, voice mode, image generation, study mode, quiet hours and whether the teen's conversations are used to improve models.

• Parents may be notified when trained reviewers identify a serious self-harm concern or behavior possibly linked to disordered eating, or when a teen account is deactivated over violence-related conversations. OpenAI says notifications do not include conversations or transcripts.
• OpenAI says parental controls do not let parents read or monitor a teen's conversations.
• OpenAI says that when ChatGPT identifies a linked teen as 18 or older, parental controls and the link are removed and the account holder gets full control of their settings. It says it notifies both the account holder and the parent, and that the account holder can set up parental controls again only by accepting a new invitation.

Privacy of age verification data

In September 2025, Altman wrote that asking adults for ID is a privacy compromise that OpenAI believes is a worthwhile tradeoff. OpenAI's help center says that for Persona verification, OpenAI does not receive the ID or selfie, only the age-related information needed to set the account's experience. Its help center article on age prediction says Persona deletes the ID or selfie within 7 days and OpenAI receives only a date of birth or an age result.

OpenAI's privacy policy says it retains personal information for only as long as it needs it to provide the service, among other stated purposes, and gives users ways to request deletion, some through the account and others by email. Our guide to ChatGPT age verification covers the verification steps in more detail.

Laws that may apply to age verification and minors

Whether ChatGPT counts as a companion chatbot or a conversational AI service under any of these laws is a legal question. These laws, bills and actions bear on how chatbot companies handle age:

• California SB 1119 (Adam's Law): authored by Senator Steve Padilla, with Assemblymembers Buffy Wicks and Rebecca Bauer-Kahan as co-authors, according to the Governor's office. The Governor signed it on September 10, 2026 (Chapter 190, Statutes of 2026). Its text says operators of companion chatbots must either determine each user's age, using age signals under the state's Digital Age Assurance Act or, if that fails, an age determination under Health and Safety Code section 27001, or apply child protections to all users. The age assurance rules take effect January 1, 2027, and most other operator duties begin July 1, 2027.
• Colorado HB 26-1263 (Conversational Artificial Intelligence Service Operator Requirements): signed May 29, 2026 (Chapter 208, Session Laws of Colorado 2026), with operator duties beginning January 1, 2027. Its sponsors are Representatives Sean Camacho and Javier Mabrey and Senators John Carson and Iman Jodeh. The act says operators must use commercially reasonable or generally accepted methods to estimate the age of account holders and users, and may not willfully disregard clear and convincing information that a user is a minor.
• The federal GUARD Act (S. 3062): introduced by Senator Josh Hawley on October 28, 2025; Senate Judiciary ordered it reported with an amendment on April 30, 2026, and Senator Grassley reported it on May 11, 2026, when it was placed on the Senate calendar (Calendar No. 406). It is a bill, not law. As reported, it would require users of AI companions to have accounts verified as adult, bar minors from AI companions, not accept a self-reported birth date alone, and limit how age-verification data is kept and shared.
• The Federal Trade Commission announced Section 6(b) orders on September 11, 2025 to seven companies, including OpenAI OpCo, LLC, seeking information on how their companion chatbots affect children and teens, including how they limit minors' use and enforce age limits. It is an information-gathering study, not an enforcement case.

We are not saying OpenAI is covered by, or out of compliance with, any of these laws. It may have practices it has not published, and whether a law applies to it is a legal question this guide does not decide.