ChatGPT Launches Age Prediction: AI Finally Learns to Treat People Differently

On January 20, 2026, OpenAI officially launched the "Age Prediction" feature in the consumer version of ChatGPT. Instead of relying on users to self - report their ages, it automatically identifies users under 18 through multi - dimensional behavioral signals such as account duration, active periods, and interaction patterns, and enables exclusive security protection. At the same time, it is accompanied by a parental control and third - party verification mechanism, marking a new stage in the protection of minors on AI platforms from "voluntary declaration" to "behavior recognition".

The Technical Logic of ChatGPT's Age Prediction

For a long time, the protection of minors on AI platforms has mostly relied on the passive model of "user self - reported age + content classification". When users check "over 18 years old" during registration, they can unlock all functions. This method is not only easy to be circumvented but also unable to handle the scenario where minors use adult accounts fraudulently.

The "Age Prediction" feature launched by OpenAI this time completely breaks this traditional logic. Its core is a multi - dimensional prediction model based on account and behavioral signals. The specific analysis dimensions include:

Account dimension: Basic information such as registration duration, account activity, and payment status;

Behavior dimension: Daily active periods (such as whether it is frequently used late at night), interaction frequency, preference for question content, dialogue length and style, etc.;

Supplementary dimension: Combine the age information filled in by users during registration, but only as an auxiliary reference, not as the sole basis for judgment.

The core advantage of this model lies in "dynamic recognition". Different from a one - time age declaration, it can continuously analyze users' usage behaviors and constantly correct the age judgment results. Even if adult users use it in a way similar to minors for a long time (such as frequently asking low - age - appropriate content or having high - frequency interactions late at night), they may be marked as "suspected minors" and trigger protection. Conversely, it is also difficult for minors to completely avoid recognition if they imitate the usage habits of adult users.

A "Combination of Hard and Soft" Solution for Minors' Protection

For accounts determined to be those of minors, ChatGPT will mandatorily enable five - layer security protection to precisely block high - risk content, specifically including:

1. Directly displayed violent and bloody images;

2. Dangerous viral challenges that may induce minors to imitate (such as extreme pranks and dangerous experiments);

3. Role - playing (Role Play) content involving sex or violence;

4. Descriptions and guidance related to self - harm and suicide;

5. Content promoting extreme aesthetics, unhealthy dieting, or body shaming.

At the same time, to avoid the impact of model misjudgment on the experience of adult users, OpenAI has introduced the third - party identity verification service Persona. Users who are misclassified as minors can complete a quick face verification by uploading a selfie. After passing the verification, they can restore the full functions of their accounts, balancing security and user experience.

In addition, the system is also equipped with a parental control customization function, giving parents more flexible control rights. They can set "quiet time" (prohibited usage periods, such as class time and sleeping time), control the account's memory function permission (to prevent children from repeatedly viewing sensitive content), and even receive timely notifications and intervene when the system detects signs of acute psychological distress in users (such as frequently asking questions related to self - harm).

Why did OpenAI launch age prediction at this time?

The launch of this feature is not an "active innovation" by OpenAI but the result of the joint promotion of regulatory pressure and industry trends.

On the one hand, OpenAI is facing an investigation by the US Federal Trade Commission (FTC). The core question is "the negative impact of AI chatbots on teenagers", and there are also multiple related lawsuits. Previously, some parents have complained that ChatGPT failed to effectively block harmful content, resulting in minors being exposed to violent and pornographic information and even having psychological problems. Launching the age prediction feature is a key measure for OpenAI to deal with regulatory reviews and reduce legal risks.

On the other hand, the protection of minors has become a "must - answer question" in the global AI industry. With the popularization of AI tools, more and more teenagers use ChatGPT as an important tool for learning and entertainment. However, their minds are not yet mature, and they are easily induced by bad information. Previously, competitors such as Google Bard and Anthropic Claude have launched different levels of minor protection functions, but most of them rely on content classification and voluntary declaration. OpenAI's "behavior recognition + dynamic protection" model is undoubtedly a more advanced exploration in the industry.

From the perspective of industry trends, the security protection of AI platforms is upgrading from "content filtering" to a dual - mode of "user recognition + content classification". It is necessary not only to judge "whether the content is harmful" but also to judge "whether the user is suitable to access the content", which is also the core direction of future AI security development.

Can age prediction truly protect minors?

Although the function design seems perfect, "age prediction" still faces many controversies and challenges, mainly concentrated in three aspects:

1. Can behavioral signals fully represent age?

The core of the age prediction model is the "correlation between behavior and age", but this correlation is not absolute. For example, some adult users may use ChatGPT frequently late at night due to work or study needs, or prefer to ask low - age - appropriate science popularization content, and are easily misjudged as minors. Some precocious minors may avoid recognition by imitating the interaction patterns of adult users. Although OpenAI says it will continuously optimize the model accuracy, it is still difficult to achieve 100% accuracy in the short term.

2. Does behavioral analysis violate user privacy?

Age prediction requires collecting and analyzing a large amount of users' behavioral data, including active periods, interaction content, usage habits, etc., which has raised users' concerns about privacy leakage. How does OpenAI ensure that these data will not be misused? Will it be shared with third parties? Although OpenAI has not clearly stated the data usage rules, in the context of stricter global data compliance, how to balance "behavior recognition" and "privacy protection" will be a problem that it must solve.

3. Can the protection cover all risk scenarios?

The five types of high - risk content blocked by ChatGPT this time mainly focus on "overtly harmful information", but the age prediction model has not covered "hidden risks" (such as inducing minors to engage in online fraud, spreading extreme ideas, and leaking personal information). In addition, the parental control function relies on the active operation of parents. If parents lack the awareness or technical ability to supervise, the actual effect of this function will be greatly reduced.

The launch of the "Age Prediction" feature on ChatGPT is an important breakthrough in the protection of minors in the AI industry. It marks that AI platforms have finally learned to "tailor services to different users", shifting from passive content filtering to active user recognition and precise protection.

However, we also need to clearly recognize that technology is not omnipotent, and age prediction is only the "first step" in the protection of minors. In the future, only when platforms, parents, and regulatory agencies cooperate, continuously optimize technology, improve rules, and strengthen guidance can we truly create a safe and healthy AI usage environment for teenagers, enabling AI technology to truly empower the growth of minors rather than bringing risks.

For OpenAI, the launch of the age prediction feature is a key step in dealing with regulations and reshaping its reputation. For the entire AI industry, it is a signal of "security upgrade". Only when technological innovation and security support go hand in hand can AI truly move towards maturity and compliance.

This article is from the WeChat official account "Shanzi", author: Rayking629, published by 36Kr with authorization.