OpenAI launches Age Prediction to flag under 18 users

If you are like me and use ChatGPT extensively, you may have already experienced the product becoming more personalized. More memory is being utilized. More shortcuts are being added. More instances where it seems to pre-empt your needs. Now, there is another type of anticipation showing up, and it has nothing to do with your preferences.

OpenAI explains that ChatGPT is introducing an age prediction model, which guesses whether an account belongs to someone under 18. If the model guesses under 18, ChatGPT automatically enforces safety measures. If you're an adult and wrongly caught by these measures, OpenAI states you can prove your age using the ‘Persona’ feature to have them lifted.

This is being marketed as teen safety, which is true. But this is exactly the kind of gate you build if the product is going to revolutionize how it makes money.

OpenAI age prediction is the new gate

OpenAI claims the system relies on signals related to your account. They might be examining general topics you talk about, the times of day you use ChatGPT, or other usage patterns.

OpenAI calls these "signals," but it is a behavioral footprint. It watches how your account conducts itself over time and uses this data to create a probability score of whether you are likely under 18.

OpenAI is also straightforward about the default: If the model determines you are under 18, it enables additional safety features. If it makes a mistake, you can verify.

ChatGPT advertising ensures age gating is inevitable

OpenAI also describes its plan to test ads in ChatGPT with these guardrails:

Ads will be separated from responses.

They will not be shown in accounts where a user claims to be under 18 or is predicted to be under 18.

Ads will not appear near "sensitive or regulated topics" such as health, mental health, and politics.

That is the clue. You cannot run a clean ad business alongside messy content, and you cannot run an ad business at all if you don't understand the audience you are serving.

Adult mode is why the line has to be sharp

It’s not a rumor. This was a direct quote from Sam Altman on X: “As age-gating continues to roll out more fully, OpenAI plans to allow even more, like erotica for verified adults.”

That's the real reason the gate must be sharp. You can't weaken the model for adults, the way people have been clamoring for, without risking headlines the first time a minor ends up in the same lane.

The trade that you will actually feel

The first thing you’ll notice is friction. Some adults will get pushed into a more restricted experience by error. OpenAI’s response is verification through "Persona."

Persona deletes your ID or selfie in seven days. But OpenAI retains what is important: your date of birth and your verification status.

That is a trade-off like no other. You provide proof once, and OpenAI guarantees no more guessing games, as it will stop running age prediction on your account.

What does this mean for work automation?

If you use ChatGPT for work, this won't stay confined to teenage safety. Practically, it brings a new failure mode that manifests at the worst possible time.

A false positive doesn't just make you slower; it breaks the flow. If you work at late-night session like me. And it looks like user behavior under 18 to ChatGPT, your experience may switch to a harsher mode or your prompt may be blocked, and the whole process freezes exactly when it needs to keep moving.

Once the product begins gating content based on who it believes you are, it rarely stops at one reason. Those gates keep opening up to whatever the platform wants to regulate next.

The ending you should keep in mind

This is not really about whether ChatGPT can correctly infer your age. It is about whether ChatGPT can determine which ChatGPT you get to use.

Ads promote cleaner segmentation. Adult mode promotes stricter eligibility. Age prediction is how this becomes achievable without every expansion being about kids.

The age of the universal chatbot is coming to a close. We are entering the age of a tiered system, where what you can access is determined by what the system believes you are and what you are willing to verify.