A university student’s cautionary note on AI in classrooms
When I started university in 2024, my professors were adamant about keeping artificial intelligence out of the classroom. This stance seemed reasonable to me. I was at university to learn, and outsourcing my papers and translations to AI didn’t seem like a great way to achieve this.
So, I was caught off guard when I saw the federal government’s recent consultation to support its “ambitious” adoption goals and plan to integrate generative AI into primary, middle, and high schools, covering kindergarten to grade 12.
Even as someone who grew up with the internet, AI often feels foreign. I barely trust it to write an email for me let alone help teach the next generation. With its rapid development, long-term studies aren’t available, and emerging research points to a number of risks for children and youth.
The government needs to back its ambitions for AI adoption with dedicated research funding to understand and address the social, educational, and emotional risks of generative AI. Without this foundation, AI doesn’t belong in K-12 classrooms.
Risks to foundational skills and social development
If existing AI literacy kits are any indication, AI curriculums will begin while children are still gaining crucial social and emotional skills. Emerging research shows how chatbots’ excessively agreeable outputs discourage social behaviour, potentially undercutting this development.
The Alberta Teachers Association's May resolution to keep “anthropomorphic AI tools” out of K-12 classrooms aligns with these findings, and illustrates that educators recognize the dangers of early AI integration.
From a learning perspective, AI’s ability to generate text and summarize passages could leave students without foundational analytical and literacy skills. Learning to write an email or plan a debate fosters important critical thinking and communication skills, and engaging with texts develops empathy and students’ sense of identity.
Productive workers begin as capable students. Until we can firmly position AI as a tool rather than a replacement for learning, introducing it into classrooms risks leaving students dependent on it for basic competencies.
Equity concerns
All chatbots are trained on data, and data has biases. Academics and journalists have already shown that chatbots perpetuate and exaggerate real-world racial and gender disparities. Putting AI in classrooms would perpetuate harmful social paradigms and jeopardize minority students.
Stanford’ University’s Faye-Marie Vassel found that AI-generated narratives are “hundreds to thousands of times” more likely to portray minority characters in stereotyped and subordinated contexts than white characters. Although this study is American, the research highlights the dangers of introducing AI into teaching for students from minority backgrounds. A Canadian study using a similar methodology has not been published as of July 2026.
AI-generated images are no better. In a 2023 Bloomberg article on the image-generating chatbot Stable Diffusion, Leonardo Nicoletti and Dina Bass wrote that “the world according to Stable Diffusion is run by white male CEOs. Women are rarely doctors, lawyers, or judges. Men with dark skin commit crimes while women with dark skin flip burgers.”
Without proper regulation, AI programs may inadvertently push misogyny, racism, and hate into classrooms despite the best efforts of teachers, staff and parents.
Will Canada fall behind?
With Canada’s productivity lagging behind our G7 peers, many are looking to AI as a silver bullet. Although it may be a useful tool for professionals, the social and emotional risks to young people should make us hesitate before making an AI tool a standard part of the curriculum.
Gen Z has already grown up encased in a hostile media landscape that sets us up for long-term challenges later in life. One in four youth aged 12–17 have reported experiencing cyberbullying, leaving them at an increased chance of poor mental health outcomes and suicidal thoughts.
While it’s good to see policy makers acknowledging the negative impacts of social media on young people, attempts at regulation mean little if we push the next generation into yet another unsafe situation.
Building digital sovereignty must include a plan to protect young Canadians from dangers posed by AI technologies. If it passes, Bill C-34, the Safe Social Media Act, would mandate tech companies “implement design features to make their [AI chatbots] safer for children.”
This is a good start, but if we want to truly safeguard our children and youth, policymakers must act on existing calls for targeted regulation of technology in education.
These regulations must be grounded in evidence. Although June’s AI strategy committed $50-million to “expand the capabilities of the Canadian AI Safety Institute” and “conduct transparent evaluations of AI models,” dedicated funding for safety research on AI in schools remains absent. Government must fund and collaborate with Canada's research institutions to inform legislation with detailed multi-disciplinary studies tailored to regional contexts.
A nation that gambles with its young people puts its very existence on the line. We cannot allow a whole generation of Canadians to be treated as collateral in the pursuit of national “productivity gains.”
Perhaps one day AI will feel as familiar as the laptop I am using to type this article. But until we have a better idea of the long-term social, emotional, and educational impacts of AI use, we owe it to our children and young people to proceed with caution, lest we leave yet another generation’s well-being in the hands of for-profit tech companies.
Aliyah Daya is a student at the University of King’s College in Halifax, N.S., where she studies both classics and sustainability.
The Hill Times