Manitoba Wants to Ban Kids From AI Chatbots. The Kids Have Thoughts.

Manitoba Premier Wab Kinew announced this week that the province will legally prohibit minors from using social media and AI chatbots, citing values that he said the platforms don't share.[¹] The announcement landed not in a policy vacuum but in an already-charged conversation — and the most pointed response came not from civil liberties groups or tech lobbyists, but from the teenagers the policy is designed to protect. "Not everything deserves to be banned," kids told CBC News, in a line that captures exactly how paternalistic internet regulation tends to collide with the people it's meant to shield.[²]

What makes Manitoba's move distinct from the chorus of similar proposals circulating in Australia, the US, and the UK is the scope. Most youth digital safety legislation targets social media platforms specifically — the algorithmic feeds, the engagement loops, the documented harms to adolescent mental health. Manitoba's bill brackets AI chatbots alongside TikTok and Instagram, treating them as a unified category of digital threat. That framing is doing real conceptual work, and it's worth scrutinizing: the harms from a recommendation algorithm optimized for outrage are structurally different from the harms of a conversational AI tool, and lumping them together may make both harder to address meaningfully. The province hasn't yet specified how it plans to enforce the prohibition,[³] which is the tell — age verification at scale remains an unsolved problem everywhere it's been attempted.

The response on Bluesky split predictably along generational and ideological lines, but one observation cut through: a commenter drew a distinction between a university education — which ideally involves being challenged, unsettled, made to unlearn — and the kind of confident-sounding knowledge people absorb from AI chat or algorithmic feeds, both of which are designed to confirm rather than complicate.[⁴] It's a sharper framing than most of the policy debate manages. The question isn't whether teenagers should use AI; it's whether the specific design of these tools — sycophantic, frictionless, optimized for engagement — develops or erodes the capacity for critical thought. Manitoba's blunt prohibition sidesteps that question entirely.

Running parallel to the youth-access debate is a grimmer undercurrent that no legislation has yet seriously addressed: the slow AI and misinformation contamination of social platforms themselves. One commenter's estimate that half of Facebook posts are now generated by AI accounts or bots[⁵] is almost certainly imprecise, but the directional anxiety it reflects is consistent with what researchers and platform engineers have been observing for two years. A related thread captured the broader fear well: the web's accumulated human knowledge — the searchable, linkable, argued-over record of what people actually thought — is being diluted by machine-generated content at a rate that makes the original seem, in retrospect, like a finite and irreplaceable resource. The AI colonization of social feeds didn't happen suddenly; it happened the way most platform degradation does, incrementally and then all at once.

Manitoba will almost certainly not be the last province, or the last jurisdiction, to attempt this kind of sweeping prohibition — and the RCMP and federal government's skepticism suggests the legal architecture isn't ready for the political ambition.[⁶] But the more durable story here isn't about whether the ban will work. It's that governments have decided the default question is no longer "how do we design these tools responsibly" but "should children be allowed near them at all." That's a significant retreat from the techno-optimist framing that dominated five years ago — and the teenagers pushing back aren't wrong to notice that nobody asked them.