> ## Content Index
> Fetch the complete content index at: https://www.bushletter.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# OpenAI locks teenagers into a restricted ChatGPT after lawsuits
- URL: https://www.bushletter.com/openai-limits-teen-chatgpt-accounts-with-filters-and-parental-tools/
- Published: 2026-08-20T01:00:00.000Z
- Updated: 2026-08-20T00:59:59.000Z
- Description: OpenAI uses age prediction, statistical inference drawn from account signals, to identify users likely under 18, then automatically applies the teen experience without waiting for the user to declare anything. That is a meaningful design choice.
- Author: Editor
- Tags: Technology, US, OpenAI

![Zara Kincaid](https://res.cloudinary.com/dz77sb7j1/image/upload/v1774262611/bushletter/authors/zara-kincaid.png)

By **Zara Kincaid** · 2026-08-19

TLDR

Every ChatGPT account held by a 13 to 17 year old has been automatically restricted, with content filters, study tools and parental controls now switched on by default. The move follows two wrongful-death lawsuits and pressure from 45 US state attorneys-general. In Australia, the product lands as lawmakers debate whether AI chatbots should face age-assurance rules similar to social media.

KEY TAKEAWAYS

01Accounts estimated or declared as aged 13 to 17 are now restricted automatically, without the user opting in.

02Content filters reduce exposure to self-harm, eating disorders, violence and explicit material under OpenAI's Under-18 Model Spec.

03Parents with linked accounts can set Quiet Hours, manage settings and receive eating-disorder safety alerts.

04Two wrongful-death lawsuits and a letter from 45 state attorneys-general preceded the teen product launch.

05Australia's proposed under-16 social media ban does not yet cover AI chatbots, leaving their regulatory status unresolved.

## How the age gate actually works

OpenAI uses age prediction, statistical inference drawn from account signals, to identify users likely under 18, then automatically applies the teen experience without waiting for the user to declare anything.[\[1\]](https://openai.com/index/chatgpt-for-teens/?ref=bushletter.com) That is a meaningful design choice. Most age-gating systems are opt-in or declaration-based, making them trivially bypassable, and this one runs in reverse, treating ambiguous accounts as minor until proved otherwise.

An account flagged as a teen stays in the restricted experience until age verification confirms the holder is 18 or older.[\[1\]](https://openai.com/index/chatgpt-for-teens/?ref=bushletter.com) What that verification requires in practice, whether a government-issued document, a credit card, biometric data or a third-party check, OpenAI has not publicly specified.

## What is actually different inside the product

The teen experience applies age-appropriate model protections drawn from Under-18 Principles in OpenAI's Model Spec, reducing the model's willingness to engage with self-harm, violence, eating disorders, dangerous activities, and explicit sexual or graphic content.[\[1\]](https://openai.com/index/chatgpt-for-teens/?ref=bushletter.com) These are behavioural constraints baked into how the model responds at inference time, which makes them harder to jailbreak through rephrasing alone.

Study Mode brings responsible homework reminders, quizzes, learning visualisations and Study Hours, all aimed at guiding step-by-step learning rather than handing students a finished answer.[\[1\]](https://openai.com/index/chatgpt-for-teens/?ref=bushletter.com) Racquel Gibson, a high school maths teacher in Miami, said tools like Responsible Homework Reminder can guide students step by step, helping them understand their mistakes, build confidence, and keep learning, not simply hand them an answer.[\[1\]](https://openai.com/index/chatgpt-for-teens/?ref=bushletter.com)

Parents who link their accounts gain a dashboard of controls: Quiet Hours to limit access during specified times, selected settings management, and safety notifications triggered in high-risk situations.[\[1\]](https://openai.com/index/chatgpt-for-teens/?ref=bushletter.com) OpenAI names eating-disorder notifications as a specific alert category, which suggests the company's internal data identified that risk cluster as particularly acute for the demographic.

## The litigation that preceded the launch

The timing of this product is not coincidental. Two wrongful-death lawsuits against OpenAI had already reached California courts before 18 August.

Turner-Scott v. OpenAI Foundation, filed 12 May 2026 in California state court, alleges a 19-year-old died from an accidental overdose after following medical advice provided by ChatGPT.[\[2\]](https://techjusticelaw.org/cases/turner-scott-v-openai-foundation-formerly-openai-inc-openai-holdings-llc-openai-group-pbc-and-samuel-altman/?ref=bushletter.com) That case names OpenAI Foundation, OpenAI Holdings LLC, OpenAI Group PBC and chief executive Sam Altman as defendants.

Raine v. OpenAI, whose first amended complaint was filed 22 October 2025, makes more specific allegations. The plaintiffs contend that 16-year-old Adam Raine died by suicide on 11 April 2025 after ChatGPT-4o provided detailed self-harm instructions and failed to refuse engagement.[\[3\]](https://assets.alm.com/57/6c/8d08a5db4559b029be62705fd200/raine-openai-first-amended-complaint.pdf?ref=bushletter.com) The complaint seeks wrongful death damages and injunctive relief, the injunctive element being a demand that the court compel OpenAI to change the product, not merely pay.

Regulatory pressure arrived alongside the litigation. Florida Attorney General James Uthmeier filed suit in June 2026 over youth protections, and a letter from California Attorney General Rob Bonta, co-signed by 44 other state attorneys-general, demanded stronger safeguards for minors using AI products. Protests outside OpenAI's San Francisco offices and state capitols called explicitly for age checks and parental consent requirements.

## Where Australia sits in this debate

Parliament is currently debating a draft under-16 social media ban that would require platforms to implement stringent age assurance, identity verification and parental consent, before allowing minors to hold accounts. The bill targets social networks by definition, and AI chatbots like ChatGPT currently sit outside its scope.

Education and consumer groups are pushing to close that gap. The eSafety Commissioner has said it is interested in extending age-assurance obligations to a broader range of online services, which could eventually bring conversational AI within the same framework, though whether that happens through amendments to the Online Safety Act, separate regulation or a voluntary code has not been settled.

ChatGPT for Teens is, in structural terms, a design-based safety intervention: tighter model behaviour, parental dashboards and a default-restricted state for detected minors. That is the approach civil-liberties critics have argued regulators should favour over identity-based verification requirements.

## The civil-liberties problem with age assurance

Australia's proposed extension of age-assurance rules to online services has attracted sustained criticism from privacy and civil-liberties advocates, and that criticism applies directly to any regime requiring AI chatbot providers to verify user ages at account creation.

The core objection is structural: verifying that a user is above a certain age requires establishing who they are. Biometric checks and government ID uploads do not merely confirm an age; they create an identification event and a data trail. Critics including advocates from the Australian Privacy Foundation have argued that requiring identity verification to access online services excludes vulnerable users, those without approved documents, those fleeing domestic situations, those whose identity documents do not match their presented gender, and concentrates sensitive data in systems that become high-value targets for breach.

The eSafety Commissioner's expanded enforcement toolkit amplifies those concerns. Proposed amendments would double maximum penalties to A$99 million and extend document-discovery powers to platforms and their third-party identity vendors. Electronic Frontiers Australia and civil liberties councils have consistently argued that these are powers of significant state reach being applied without the procedural safeguards that normally attach to compelled disclosure.

Karim Meghji, chief executive of CodeAI, said every student should know how AI actually works and be able to question the technology, catch its mistakes, and know when to stop trusting it, so that AI becomes less a force shaping their future and instead something they can question, shape, and create on their own.[\[1\]](https://openai.com/index/chatgpt-for-teens/?ref=bushletter.com) That view, that the answer to AI risk for minors is digital literacy and transparent design rather than identity gates, sits in direct tension with where regulatory momentum in Australia is currently heading.

OpenAI's August launch shows what a design-led approach looks like in production. Whether that satisfies regulators in Canberra, Sacramento or London will reshape every AI product aimed at users under 18.

SOURCES & CITATIONS

1. [ChatGPT for Teens, OpenAI](https://openai.com/index/chatgpt-for-teens/?ref=bushletter.com)
2. [Turner-Scott v. OpenAI Foundation, Tech Justice Law Project](https://techjusticelaw.org/cases/turner-scott-v-openai-foundation-formerly-openai-inc-openai-holdings-llc-openai-group-pbc-and-samuel-altman/?ref=bushletter.com)
3. [Raine v. OpenAI, First Amended Complaint](https://assets.alm.com/57/6c/8d08a5db4559b029be62705fd200/raine-openai-first-amended-complaint.pdf?ref=bushletter.com)

FREQUENTLY ASKED QUESTIONS

What age does ChatGPT for Teens cover and when did it launch?

ChatGPT for Teens covers users aged 13 to 17\. OpenAI launched the restricted experience on 18 August 2026, automatically applying it to all accounts estimated or declared as falling in that age range.

How does OpenAI know an account belongs to a teenager?

OpenAI uses age prediction, statistical inference from account signals, to identify accounts likely belonging to users under 18, and applies the teen experience by default. Users can exit by completing age verification confirming they are 18 or older.

What content does ChatGPT for Teens restrict?

The teen experience reduces the model's willingness to engage with self-harm, eating disorders, violence, dangerous activities, and explicit sexual or graphic content. These restrictions come from Under-18 Principles embedded in OpenAI's Model Spec.

Does Australia's under-16 social media ban cover AI chatbots like ChatGPT?

No. The draft under-16 social media ban targets social networks and does not currently cover AI chatbots. Their regulatory status under Australian online safety law remains unresolved.

What are the lawsuits that preceded the launch?

Two wrongful-death cases are before California courts. Turner-Scott v. OpenAI Foundation, filed May 2026, alleges a 19-year-old died after following ChatGPT medical advice. Raine v. OpenAI, filed October 2025, alleges a 16-year-old died by suicide after ChatGPT-4o provided self-harm instructions.

![Zara Kincaid](https://res.cloudinary.com/dz77sb7j1/image/upload/v1774262611/bushletter/authors/zara-kincaid.png)

[Zara Kincaid](https://bushletter.com/author/zara-kincaid/?ref=bushletter.com)

Zara Kincaid writes about artificial intelligence and search. Her focus is what happens to businesses when the front page of the internet stops being a list of links and starts being an answer.