If you're a parent and your teenager does homework, applies to college, or apparently just exists, they've almost certainly used ChatGPT. The minimum age since day one has been 13. The content restrictions for 13-year-olds, until this week, were exactly the same as for everyone else.

OpenAI launched ChatGPT for Teens on August 18, a dedicated mode for users aged 13 to 17 with built-in content restrictions and optional parental controls. The mode is real and the protections are meaningful. It's just arriving in 2026, roughly three years after teenagers decided ChatGPT was the right tool for biology papers, college essays, midnight anxiety spirals, and everything in between.

What actually changed

Four things are different in teen mode.

Content restrictions. Teen mode limits how ChatGPT responds to topics around self-harm, suicide, eating disorders, violence, and sexual content. Not a hard block — more like a posture shift. A teen asking about a friend who mentioned self-harm gets a redirected response with resources rather than a clinical walkthrough. The difference is real. It's not a content filter that pretends to work and doesn't.

Study Mode. This is the one that might actually matter day to day. Teen mode defaults to a homework-help posture that gives guiding questions and step-by-step support rather than just finished answers. The idea is teaching, not completion. Whether that works in practice depends on how determined your teenager is to get around it, which is to say: it works until it doesn't.

Parental controls. Parents can set Quiet Hours (ChatGPT unavailable during those windows), Study Hours (Study Mode locked on), and opt in to alerts when OpenAI's safety systems flag an at-risk conversation. These require a linked parent account, which is opt-in. That means they only work if you actively set them up. Most parents won't. OpenAI knows this.

Age-prediction. OpenAI says ChatGPT will route users to teen mode when its systems estimate they're 13–17, even if they've never declared their age. The method behind this is not publicly disclosed.

Source spread

• OpenAI — ChatGPT for Teens [hype] — company's own framing; emphasizes "stronger built-in safety protections," Study Mode, and parental controls.
• TechCrunch [skeptic] — headline leads with "years after teens started using it"; covers the gap between when teens adopted ChatGPT and when protections arrived.
• CNBC [safety] — covers the safety context and ongoing child-safety scrutiny around AI platforms.
• CNN Business [skeptic] — reports launch with reference to broader questions about AI's effects on teen mental health.

What's real and what deserves a side-eye

What's real:

• The content restrictions are meaningful, not cosmetic. What teen mode returns for eating-disorder questions, self-harm prompts, or violent content is genuinely different from what standard ChatGPT returns. This is real.
• Study Mode as a default is a good instinct. The model asking "what do you think the answer is?" before walking through a problem is how homework help is supposed to work.
• Parental alert notifications for flagged conversations are worth enabling if you have a teenager who's going through something hard. They're not surveillance — they're a smoke alarm.

What deserves a side-eye:

• The age-prediction method is black-box. A teenager who lies about their age gets standard ChatGPT. That cost has always been zero. It still is.
• Parental controls require active setup. The controls work well once enabled. Most families won't enable them. The product is designed so that opt-in failure is the default outcome.
• These guardrails are for the incoming cohort. Teenagers who've been using unmodified ChatGPT since 2023 are now 16, 17, 18. The mode addresses the next generation of teen ChatGPT users, not the ones who grew up on the unguarded version.

What to do about it

• Link a parent account if you have a teenager. The content restrictions activate automatically, but parental controls — Quiet Hours, risk alerts — require a connected account. Takes about ten minutes. Instructions are in ChatGPT settings.
• Set Quiet Hours during sleep. Late-night ChatGPT use is the most documented risk pattern for teen mental health. Quiet Hours exist precisely for this. Pick a window and enable it.
• Lock Study Hours during homework time. If your kid has a set study period, locking Study Mode on during those hours means ChatGPT defaults to guiding questions rather than finished answers.
• Don't assume age-prediction catches everything. A teenager who enters a birth year that makes them 21 gets standard ChatGPT. Have an actual conversation with your kid about what to use it for and what the limits are.
• For important health questions, verify elsewhere. Teen mode is better than standard ChatGPT for mental health topics. But "better" doesn't mean "substitute for a doctor or counselor." Cross-check anything health-related with a real professional.

Further reading

• OpenAI — ChatGPT for Teens announcement — the full list of features, parental controls, and Study Mode details
• TechCrunch — OpenAI launches a safer ChatGPT for teens — useful framing on the timing gap
• CNBC — OpenAI rolls out ChatGPT for Teens — child safety context
• NBC News — ChatGPT teen safety measures — covers age-prediction and what parents can't do