The Misinformation Conversation Is Stopping Being Scared and Starting Getting Practical

The alarm has been running for months. Deepfakes targeting Kamala Harris, AI-generated attack ads threatening electoral integrity, a watchdog invoking Taylor Swift to make the point that synthetic media is a democratic hazard — the AI-misinformation conversation has been saturated with a particular flavor of dread. But in the last 24 hours, something in the register shifted. The dominant mood across platforms moved from fearful to pragmatic, not because the threats receded but because the conversation found a new object to organize around: rules. Brazil announced that candidates who use AI to spread disinformation could lose their mandates. Manitoba introduced an election misinformation bill. Suddenly the discourse had legislative scaffolding to attach itself to, and the energy changed.

The platform dynamics are telling here. Bluesky, which skews toward researchers and media-adjacent professionals, is running the most analytically grounded content — including a circulated study examining how news users in Mexico, the US, and the UK navigate AI-generated misinformation across different epistemic cultures. That's the community trying to understand the problem structurally, not just react to it. YouTube commenters, by contrast, are still in the most alarmed register of any platform, with the most negative average sentiment of the group. News coverage is doing the institutional work of cataloguing the threat — deepfake trends in 2024, the Grok backlash prompting congressional calls for a crackdown — but it's also, as the Columbia Journalism Review piece on Cuomo noted with something approaching dark humor, documenting how accessible all of this has become. You can deepfake yourself now. Anyone can.

What remains striking is what the pragmatic turn doesn't resolve. The deepfake abuse conversation — specifically, that roughly 90% of deepfakes circulating online are non-consensual sexual content targeting women and girls — has been gaining real traction on Bluesky, with advocates pushing back against media coverage they see as sensationalizing the phenomenon rather than centering the people it harms. This is a different kind of pragmatism from the legislative kind: not rule-making, but a demand to redirect attention. The two pragmatisms don't always point in the same direction. Electoral integrity legislation and survivor-centered advocacy share a diagnosis but diverge on priority, and the conversation hasn't yet found a way to hold both at once. What the shift away from fear reveals is that the public is ready to act — it just hasn't settled on what acting should look like.