ChatGPT for teens arrives: OpenAI's school push collides with EU privacy rules

A product launch, not a model upgrade

ChatGPT for teens is exactly what the name says: a reconfigured ChatGPT experience aimed at students roughly between 13 and 18. According to The Decoder, the practical additions matter most. Students can photograph a handwritten assignment and get step-by-step guidance instead of a ready-made answer; they can work with uploaded documents; and the safety layer is deliberately stricter than the consumer version adults get.

OpenAI also brought in Common Sense Media, the well-known US non-profit, to help shape the experience and curate educational content. That's smart optics, because the adoption bottleneck in education was never model quality. It's trust — from parents, teachers and data protection officers.

Notice that OpenAI positioned the launch as an experience change, not a model upgrade — no new weights, no new benchmark claims. We test models in AI Arena by tokens per second, time-to-first-token and VRAM usage; none of that changes with the teen version. What changes is the policy layer: what the model is allowed to say, and what happens to the data afterwards.

Where the EU friction starts: GDPR consent ages

This is where the European perspective diverges from the US press release. Under Article 8 of the GDPR, the default age of consent for processing children's personal data is 16, not 13. Member states may lower that threshold, but no lower than 13; the variation therefore comes from national decisions to lower the age, not raise it. Germany and the Netherlands keep the default at 16, while the Czech Republic has lowered it to 15. For an EU school using ChatGPT for teens, the age of every student in the room determines whether parental consent is required. One classroom can hold students with three different legal statuses. That is not a hypothetical problem; it's an hour of paperwork before the first lesson.

Consent is only the beginning. A school is a data controller under the GDPR. Processing student data with OpenAI requires a data processing agreement, a valid legal basis and, in most cases, a Data Protection Impact Assessment. And since August 2025, the EU AI Act applies to general-purpose AI models as well, adding transparency and documentation obligations. None of this bans AI in classrooms — but the compliance burden sits on the school, not on OpenAI.

What happens to a teenager's chat data

The question parents actually care about: is this conversation stored, and is it used to train the model? According to OpenAI's announcement, teen conversations are not used to train models by default. If that commitment holds up contractually in the EU — and if the school's data processing agreement mirrors it — it's the most important sentence in the launch.

From our own experience running AI services in the EU, the pattern is consistent: the regulation itself is rarely the blocker; the documentation is. Schools that can show a signed DPA, a DPIA and a clear retention policy will adopt this. Schools without that paperwork won't — no matter how good the chatbot is.

Price and practical reach

OpenAI didn't announce a separate education price for the teen experience; it runs on ChatGPT's existing tiers. The free tier is enough to test the core features. ChatGPT Plus costs $20 per month — roughly €18 at current exchange rates — and adds a larger context window and higher usage limits.

Before planning a rollout, European schools should verify that the teen experience is actually active in their country. OpenAI's launches frequently start US-centric, and education features tend to reach the EU later. The practical test is simple: open ChatGPT, check the settings and see whether the teen mode exists in your account. If it isn't there, it isn't available to your students yet.

How it stacks up against Google and the EU alternatives

OpenAI is entering a classroom that Google already rents out. Gemini is embedded in Workspace for Education, which most European schools already use for Docs and Gmail; teachers get AI as a tab in the tool they already log into, not another login to manage. Microsoft's Copilot has similar reach through Microsoft 365 Education. Neither is automatically better — but in education procurement, existing contracts and data processing agreements often beat benchmark scores.

There are European alternatives too. Mistral's open-weights models are hosted on European infrastructure, which simplifies GDPR compliance, although Mistral doesn't offer a classroom suite with management features. The decision tree for most schools is blunt: if the Google or Microsoft contract already handles data residency, staying put is easier. If the school wants genuine sovereignty, a self-hosted open model may be a cleaner path — at the cost of serious IT effort. That's the trade-off nobody puts in the press release.

The bottom line

ChatGPT for teens is a reasonable, overdue product: the same capable model, stricter guardrails, and an explicit acknowledgment that teenagers need a different default than adults. For European users, though, the launch changes nothing about the fundamental calculation — GDPR consent ages that vary by member state, AI Act obligations and data residency requirements are still the real gatekeepers.

Parents: test it on the free tier. Schools: read the data processing agreement before you read the feature list. That's not a joke about bureaucracy — it's the difference between a pilot project and a compliance headache.