OpenAI has revealed more details about how ChatGPT predicts whether a user is likely to be under 18. The age-prediction system is designed as an additional safety layer that helps the chatbot determine when an account may belong to a teenager and automatically apply stronger protections. Unlike a simple age check based only on the date of birth entered during signup, the system examines a combination of behavioral and account-level signals.
The disclosure comes as OpenAI expands its efforts to make ChatGPT safer for younger users. The company says the system is not intended to determine a user’s exact age. Instead, it estimates whether an account is likely to belong to someone under 18. If the system is uncertain or believes an account may belong to a minor, ChatGPT can default to a safer experience with additional restrictions around sensitive content.
How ChatGPT Predicts Your Age
OpenAI says ChatGPT uses an age-prediction model that examines multiple signals associated with an account.
The system does not depend on one particular conversation or a single piece of information. Instead, it looks at patterns that develop over time.
Some of the signals OpenAI has identified include:
| Signal | What it can indicate |
|---|---|
| Stated age | Age information provided by the user |
| Account age | How long the ChatGPT account has existed |
| Usage times | Typical times of day when the account is active |
| Usage patterns | How and when the account is generally used |
| General topics | Subjects commonly discussed with ChatGPT |
| Activity over time | Longer-term behavioral patterns |
OpenAI describes these as behavioral and account-level signals. The company says it is continuing to learn which signals improve the accuracy of its predictions as the system is deployed.
It Does Not Simply Read Your Conversations
The disclosure is important because it shows that age prediction is broader than asking ChatGPT to guess someone’s age from the way they write.
OpenAI says the system considers general topics a user discusses, along with usage patterns, account history and other account-level information. This means the prediction is based on a combination of signals rather than a single analysis of someone’s language or appearance.
The company has not publicly disclosed the exact weighting given to each signal or the precise model architecture used to produce an age prediction.
Why OpenAI Wants to Predict User Age
The main purpose of the system is teen safety.
OpenAI’s age prediction technology is designed to determine whether an account may belong to someone under 18 so that ChatGPT can provide an age-appropriate experience.
This is particularly important because users can provide a date of birth when creating an account, but OpenAI says age prediction provides an additional safety check.
An account can therefore potentially be identified as belonging to a minor even if the user previously supplied a date of birth.
When ChatGPT predicts that an account may belong to someone under 18, it can automatically enable the ChatGPT for Teens experience, provided the account is eligible.
What Happens When ChatGPT Thinks You Are Under 18?
Users whose accounts are placed into the teen experience can continue using ChatGPT, but certain content receives additional safeguards.
OpenAI says the protections can limit exposure to areas such as:
- Graphic violence or gore
- Risky viral challenges
- Sexual or romantic roleplay
- Violent roleplay
- Self-harm content
- Content promoting extreme beauty standards
- Unhealthy dieting or body-shaming material
The goal is not to prevent teenagers from using ChatGPT altogether. Instead, OpenAI says the system is intended to provide stronger protections around content that could pose particular risks to younger users.
Adults Can Be Misclassified
OpenAI acknowledges that age prediction is not perfect.
An adult may sometimes be incorrectly identified as being under 18. In such cases, OpenAI provides an age-verification process for eligible accounts.
Users can check the age category associated with their account through Settings > Account > Age.
If an adult has been placed into the teen experience incorrectly and age verification is available, they can use the verification process to confirm that they are 18 or older.
How Age Verification Works
OpenAI uses Persona, a third-party identity-verification company, for users who need to confirm their age.
Depending on the user’s country, Persona may request a live selfie, a government-issued identity document, or both.
The verification process can check the date of birth on an identity document and, where applicable, compare the user’s selfie with the photograph on the document. The exact documents accepted can vary by country.
OpenAI says it does not receive the government ID or selfie submitted to Persona. Instead, OpenAI receives only the information needed to determine the appropriate ChatGPT experience, such as confirmation that the user is 18 or older and relevant age information.
Persona says it deletes the uploaded ID or selfie within seven days after verification.
What About Privacy?
Age prediction naturally raises privacy questions because the system uses information associated with a user’s account.
OpenAI says age prediction uses signals such as usage patterns, account history, typical usage times and general topics discussed with ChatGPT. The company says this information is used to determine the appropriate age experience and safety protections.
OpenAI also says users who verify their age through Persona do not have their ID or selfie shared with OpenAI.
The company says that after a user verifies their age, it will not continue running age prediction on that account.
Users who have questions or objections concerning age prediction can contact OpenAI through its privacy support channel.
Why the System Is Becoming More Important
The rollout of age prediction comes as AI companies face increasing pressure to protect younger users.
OpenAI has been expanding its teen-focused safeguards, including parental controls and additional protections around sensitive content. The company has also said that parents can customize certain aspects of a teen’s ChatGPT experience, including quiet hours and selected feature settings.
The issue has become more prominent as AI chatbots move beyond simple question-and-answer tools and become increasingly integrated into education, entertainment, personal advice and everyday activities.
OpenAI said in October 2026 that teenagers using ChatGPT spend less than 15 minutes per day on average, while fewer than 2% use the service for more than three consecutive hours. The company has also highlighted its efforts to add break reminders and other safety features for younger users.
The Challenge of Getting Age Prediction Right
The central challenge is balancing safety with accuracy.
If an AI system is too aggressive, legitimate adult users could face unnecessary restrictions. If it is too permissive, teenagers could receive an experience that does not provide sufficient safeguards.
OpenAI says it defaults to a safer experience when it is not confident about someone’s age or has incomplete information. The company is also continuing to refine the system based on what it learns during rollout.
That makes age prediction an evolving technology rather than a definitive age-identification system.
Age Prediction Is Not the Same as Knowing Your Exact Age
One important distinction is that ChatGPT is not necessarily determining whether a person is exactly 17, 24 or 35.
The system’s key objective is to estimate whether an account likely belongs to someone under 18. In other words, it is primarily an age-category prediction designed for safety rather than a precise measurement of chronological age.
This distinction also explains why OpenAI provides a separate verification mechanism for adults who are incorrectly categorized.
The Bigger Picture
OpenAI’s age-prediction system represents a broader shift in how AI platforms are approaching online safety. Instead of relying entirely on information users provide during signup, AI companies are increasingly using behavioral signals and account-level patterns to determine which protections should apply.
For ChatGPT, the immediate goal is to protect younger users from potentially harmful content while allowing adults to use the platform with fewer restrictions. But the approach also creates an important privacy and accuracy trade-off because the system needs to analyze signals about how accounts are used.
As AI becomes more deeply integrated into people’s daily lives, the ability to establish appropriate age boundaries will become increasingly important. OpenAI’s decision to disclose the broad categories of signals it uses gives users a clearer picture of how the system works, although the company has not revealed the precise formula or model weighting behind its predictions.
Looking Ahead
OpenAI is expected to continue refining its age-prediction system as it gathers more information about which behavioral and account-level signals provide reliable results. The company has said that accuracy improvements will be an ongoing part of the rollout, while uncertain cases will generally receive the safer experience.
The bigger test will be whether OpenAI can reduce the risk of minors accessing inappropriate content without creating excessive restrictions for adults. As ChatGPT for Teens expands and AI platforms face greater scrutiny over child safety, age prediction could become an increasingly important part of how online AI services determine which experiences users receive.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
