The Founders

The students at The Study are not an exception. They are part of a generation that is building with AI at a pace no previous generation could match.

Alby Churven is 14 years old and lives in Sydney, Australia. He started building AI startups after a video he made about Y Combinator went viral on social media.^[6] He is not old enough to drive. Under the ban, he would not be old enough to use ChatGPT.

Nick Dobroshinsky is 15. His company, BeyondSPX, uses AI to generate stock analysis reports. It has 50,000 monthly users.^[7]

Pranjali Awasthi founded Delv.AI when she was 16. The company uses artificial intelligence to transform how people do data research.^[8]

Raghav Arora co-founded GetASAP at 16. The company now has 48 employees and has raised $3.4 million in funding.^[9]

These are not edge cases. AI startups led by founders under 27 collectively raised over $100 million in 2025.^[10] Venture capital firms, including Y Combinator, no longer have strict age minimums. The barrier to building real software has dropped so far that a teenager with a laptop and an AI tool can ship a product that competes with teams of adults.

A ban treats these founders as children who need protection from technology. The venture capital market treats them as founders who deserve funding. Both cannot be right.

School lockers. Behind these walls, a generation of builders is using AI to start companies, solve problems, and create tools. A ban would lock them out of the platforms with safety features and push them toward the ones without. Photo: Public domain, via Wikimedia Commons.

The Rights

The debate about banning AI for children is also a debate about rights. Canadian law already recognizes that young people are not a single category.

Michael Geist, a professor of law at the University of Ottawa, testified before the Senate Social Affairs Committee on May 6, 2026.^[11] He argued that banning AI chatbots for children raises serious constitutional questions. Section 2(b) of the Canadian Charter of Rights and Freedoms^[12] protects the right to receive and express ideas. A teenager researching a health condition, learning to code, exploring identity questions, or doing homework with AI assistance is exercising that right.

Geist wrote: "Children are increasingly recognized as rights-bearers under the Charter and under international instruments such as the United Nations Committee on the Rights of the Child's General Comment 25 on children's rights in digital environments, and a wholesale denial of their access to a major source of information and expression is incompatible with that recognition."^[13]

General Comment No. 25, adopted by the UN Committee on the Rights of the Child in 2021, explicitly extends children's rights into digital spaces.^[14] These include the right to access information, the right to education, the right to participate, and the right to play. A ban on AI chatbots restricts all four.

The UN's "Our Digital World, Our Say" initiative, launched in 2025 by the Office of the High Commissioner for Human Rights, goes further. It asks children directly whether digital tools respect their rights.^[15] The premise is that children have standing to participate in decisions about technology that affects them. A ban made without consulting them contradicts that premise.

Canadian law already has a framework for recognizing that young people mature at different rates. It is called the mature minor doctrine.^[16] In 2009, the Supreme Court of Canada ruled in A.C. v. Manitoba that a minor below the age of majority can consent to medical treatment if they demonstrate sufficient understanding and maturity. The decision is based on capacity, not birthdate.

Bill C-27, Canada's proposed privacy legislation, extends this principle to data.^[17] It states that minors who are capable of understanding the consequences of data collection can provide meaningful consent. Privacy scholars have proposed graduated consent frameworks^[18] that adjust obligations as children mature, rather than drawing a single line at age 16.

The legal infrastructure for capacity-based access already exists. A binary age gate on AI chatbots ignores all of it.

The Substitution

The strongest practical argument against a ban comes from Geist's own testimony. A 15-year-old who cannot access ChatGPT does not stop using AI. They find something else.^[19]

The "something else" is the problem. Commercial AI platforms like ChatGPT, Claude, and Gemini are working on dedicated trust and safety teams. They have content moderation. They have crisis routing that tries to connect users expressing suicidal thoughts to help. They have age-appropriate design features, however imperfect.

A banned teenager moves to offshore AI services with no safety infrastructure. Or they use a VPN to access the banned platforms anyway. Australia's under-16 social media ban, implemented in late 2025, achieved roughly 70 percent compliance after three months.^[20] Banning AI tools would face even steeper challenges, because the same models are embedded in search engines, word processors, and operating systems that no ban can practically reach.

The ban does not remove AI from children's lives. It removes some of the safer AIs from children's lives.

"You have not protected the child. You have pushed them somewhere worse."
– Alexander Somma

An iron gate to a winding path at Vibble, Gotland. Some schools built their own way through. They own the gate. Photo: CC BY-SA 4.0, via Wikimedia Commons.

The Alternative

The choice is not between banning AI and giving children unregulated access. There is a third option.

Geist proposes an AI Transparency Act that forces companies to disclose their safety policies, training data practices, law enforcement reporting standards, and age restrictions.^[21] Not a ban. Accountability. Combined with graduated design obligations^[22] that match safety features to demonstrated maturity rather than birthdate, and real penalties for violations.

But transparency and regulation address what happens on commercial platforms. They do not help schools that want to control the environment themselves.

Some schools have built that environment. The University of Missouri's Show-Me AI initiative runs on university-owned servers under university rules.^[23] The Study in Westmount built Rosie, an AI agent on school-owned infrastructure, powered by Sage.Education's Sage.is AI-UI.^[24] Students build on Rosie's foundation using tools they choose. The school decides what runs. The school decides what is visible. The school decides where the data lives.

Sage.Education does not age-gate.^[25] It gives schools the tools to make their own decisions about access, visibility, and safety. The school writes the rules. Teachers can see how students use AI through the Learning Visibility Dashboard, and students know the teacher can see. The data stays on the school's hardware. There are no hidden classifiers, no engagement metrics, no advertising pipeline.

This is not a ban, nor is it unregulated access. It is a classroom, where the teacher is in charge, the rules are clear, and the students are builders.

Self-hosting requires technical capacity that many schools lack, and open-source platforms demand maintenance that underfunded IT departments cannot always provide. But the architecture proves that the choice between banning AI and surrendering to Big Tech's design is a false one. Schools can own the tools their students build with.

The students at The Study who built SafeTrace may replace every tool mentioned in this article before they graduate. The platforms will change, the models will evolve, the policy debates will move on to the next technology.

What will not change is their capacity to build. The question is whether we let them, or whether we ban the generation Canada is investing $2 billion to prepare.

This article is a companion to "Loyal to Whom?" which examines the design disloyalty and governance gaps that make AI chatbots dangerous in the first place. The views expressed are those of the editorial board of Sage.is AI-UI and Sage.Education, and do not necessarily reflect the positions of any institution mentioned. Sage.is AI-UI and Sage.Education are products of Startr LLC; their inclusion represents a disclosure of interest. No individuals quoted in this article were interviewed (yet); all quotes are from published sources. Full disclosure and transparency is a feature, not a bug.