A Teen Tragedy Sparks National Debate: Should Kids Be Banned From AI Chatbots?

A Teen Tragedy Sparks National Debate: Should Kids Be Banned From AI Chatbots?

The Canadian government is grappling with a monumental decision: whether to prohibit children under 16 from accessing AI chatbots. This urgent debate follows a harrowing incident in Tumbler Ridge, British Columbia, where a teen who killed eight people was found to have had disturbing discussions with OpenAI's ChatGPT before the attack. While the suspect was 18 at the time of the killings, the case has intensified public scrutiny over youth access to powerful AI tools. Culture Minister Marc Miller has stated he is "very seriously" considering such a ban, noting that regulating AI presents a more complex challenge than traditional social media.

This discussion isn't happening in a vacuum. A recent poll indicates that 69% of Canadians support restricting AI chatbots for children under 16, nearly matching the 70% who favor age restrictions for social media platforms like Instagram and TikTok. The province of Manitoba has already announced its intention to ban social media and AI chatbots for children under 16, likely starting in schools. This potential nationwide ban would position Canada as a pioneer in aggressively restricting children's access to artificial intelligence, going beyond measures like Australia's age-gate for social media. However, concerns remain about limiting access to tools that some believe offer valuable learning opportunities.

The stakes are incredibly high, pitting the potential for harm against the promise of educational advancement.

Gobble's Take: If a chatbot can be linked to a tragedy, the "educational benefits" argument suddenly feels very, very small.

Sources: The Logic · Global News · The Walled Garden Education

Khan Academy Admits Its AI Tutor Was a "Non-Event" for Most Students

After a year of watching AI unfold in education, a stark reality check has emerged from a surprising source: Khan Academy. Sal Khan, the founder of the widely respected non-profit, recently admitted that Khanmigo, their much-hyped AI tutor, was a "non-event" for a significant number of students. Despite over 108 million interactions since its 2023 rollout, only 15% of students with access to Khanmigo regularly engaged with it. This admission challenges the widespread optimism about AI's immediate impact on personalized tutoring at scale, highlighting that text-based chatbots often struggle to maintain student engagement beyond initial interactions, especially for traditional schoolwork.

The core issue isn't necessarily the AI's capability, but its integration and scalability. Current AI models often suffer from "anterograde amnesia," meaning each new session starts from scratch without learning about the student or retaining previous context. This makes truly personalized, long-term tutoring difficult and expensive. However, a glimmer of hope comes from Anthropic's Model Context Protocol (MCP), an open standard released in November 2024 that acts like a "USB-C" for AI, allowing models to securely connect to existing student data systems like learning management systems and assessment platforms. While MCP promises to simplify how AI tools access and utilize student data, its widespread adoption in EdTech is still in its early stages.

The promise of a personalized AI tutor for every student remains a distant goal, not a current reality.

Gobble's Take: Your kid isn't missing out on a revolutionary AI tutor; they're probably just happy to be done with homework.

Sources: Wess Strabelsi · ETIH EdTech News

Schools Are Rolling Out "Emotional Surveillance" and Keystroke Trackers to Monitor Students

Forget just checking homework; schools are increasingly deploying sophisticated AI systems to monitor students in ways that extend far beyond academic performance. New technologies are emerging that fall into two disturbing categories: "emotional surveillance" and "integrity surveillance." Emotional surveillance platforms use cameras to track students' attention and emotional states during screen-based work, claiming to identify everything from engagement to frustration. Meanwhile, integrity surveillance tools log keystrokes, track browser activity, and generate originality scores to detect AI use and potential cheating.

These systems, often deployed without explicit parental consent, collect a vast array of data, including biometric information and detailed behavioral assessments, which could create permanent digital profiles of students. Privacy advocates warn that this constant monitoring can lead to students self-censoring, feeling anxious, and developing mistrust, rather than fostering genuine learning and self-awareness. While proponents argue these tools are necessary for student safety and academic integrity, they often prioritize control over trust, creating an "arms race" where students develop counter-AI tactics to evade detection.

The classroom is becoming less a space for exploration and more a panopticon of constant digital observation.

Gobble's Take: Your child's school might know more about their mood swings and typing habits than you do.

Sources: The Walled Garden Education · Sohail x Codes · Cambridge Analytica

Oklahoma Now Requires "No AI" for Graduation Projects as States Pass New Classroom AI Laws

In a significant shift, states are finally beginning to establish concrete legislative frameworks for AI in K-12 education, moving beyond vague recommendations. Maryland, for instance, has passed the Artificial Intelligence Ready Schools Act (SB 720), which mandates the State Department of Education to provide comprehensive guidance on AI to schools, educators, parents, and students by June 1, 2027. This law emphasizes safe, equitable, and ethical AI use, promoting AI literacy and requiring each local school system to appoint a non-instructional coordinator to manage AI implementation.

Meanwhile, Oklahoma is taking a direct stance on academic integrity. Senate Bill 1734, known as the Oklahoma Responsible Technology in Schools Act, establishes a regulatory framework that includes parental opt-out options for student-facing AI tools and requires a "human-in-the-loop" for AI outputs used in instruction or assessment. Most notably, Oklahoma House Bill 3021 now allows school districts to waive certain graduation requirements for military family students, provided the student completes a project or paper without AI assistance. This marks one of the clearest legislative mandates against AI use in student work for graduation credit. Additionally, Tennessee's SB 677 requires the state education department to provide free professional development courses on AI usage for teachers in grades 6-12, with teachers mandated to complete at least one by August 1, 2027.

These new laws signal a growing recognition that AI in schools demands more than just guidelines – it requires enforceable rules that shape everything from curriculum to graduation.

Gobble's Take: The era of "just don't use ChatGPT" is over. Now, states are drawing hard lines that could impact your child's diploma.

Sources: Plural Policy Inc.(https://pluralpolicyinc.substack.com/p/ai-policy-watch-0528-analyzing-14) · LegiScan · BillTrack50

In Case You Missed It

Yesterday's top stories:

• School AI policies exist — but they're not landing
• AI detection doesn't catch cheating — it just moves the problem
• The real classroom question isn't which AI tool — it's who's still doing the thinking