Audiences seem to have stopped booing AI videos. That's a warning, not a win

Remember when every suspicious clip seemed to draw a wave of "ugh, AI" in the comments? Lately that reflex looks like it has faded. It would be comforting to read that as maturity, as if the public had made peace with synthetic media. We read it differently: people haven't made peace with anything, they've just run out of energy.

The boos have gone quiet, and nobody can say why

The trigger for this column is a piece from Der Spiegel, published on October 31, 2026. Its headline says that suddenly nobody seems bothered anymore that AI is everywhere on YouTube, TikTok and Instagram. The article is paywalled in the material we have, so we can't tell you what evidence it offers or what explanation it gives. We take the headline's observation as a starting point and argue about what it means.

Our view is that two things feed each other. We think detecting fakes by eye is a losing game that gets harder every month. And calling them out is a chore that earns nothing. When a feed serves you dozens of clips in a few minutes, flagging each suspicious one is no longer a habit but a job nobody hired you for.

Silence is not consent

Fewer complaints would mean acceptance only if people could still tell what they were looking at. If they can't, a quiet comment section is simply a place where nobody noticed. A drop in outrage can mean "I'm fine with this" or "I stopped checking," and from the outside the two look identical.

There is some supporting context, though it's not direct proof of our thesis. A heise report on a study says AI videos massively undermine trust in moving images, that most Germans fall for deepfakes, and that they demand strict labeling requirements. That doesn't look like a public at ease. It looks like a public that feels exposed and wants someone else to take responsibility. We'd be careful not to stretch one study into a universal law, but it points the same way.

This is also why "the market has spoken" is a poor argument. In our view, outrage was never a reliable filter; it was a stopgap that worked while fakes were clumsy and rare. Treating its decline as a verdict ignores that the filter was always volunteer labor from tired people.

The filter has to move to the platform

So who should do the work? Not the viewer. The rules are already here: legal and compliance explainers say the transparency duties in Article 50 of the EU AI Act apply from August 2, 2026, including disclosure for AI-generated content showing real people. These are undated web pages, so we wouldn't lean on their fine print. But the direction is clear enough.

The platforms already have policies to point to. One guide on social media labeling says Meta requires disclosure for photorealistic video and realistic-sounding audio made or altered with AI, with action against posts or accounts for repeated violations. A policy is not enforcement, and we'd treat the gap between the two as the real story. A rule that depends on uploaders confessing is weakest exactly where it matters, with people who want to deceive.

Our view is that three things should be non-negotiable:

• Automatic detection and provenance signals, so labels don't depend on honesty alone.
• Visible labels inside the feed, not buried in a description most people never open.
• Real penalties for repeat offenders, applied consistently and not only after a scandal.

None of this means banning AI video. Plenty of synthetic clips are harmless or even fun. The point is that the viewer should know what they're watching before deciding whether they care.

Our verdict

Don't celebrate the quiet. If audiences really have stopped complaining, that is a reason for platforms to step in, not a reason to step back. Fatigue is a resource that gets exhausted: once people stop caring whether a video is real, the next fake that matters, whether political, financial or personal, lands on an audience with its guard down. As viewers, we'd suggest a modest habit: treat unlabeled, too-perfect clips as unverified by default. As for platforms, the burden of proof should sit with them, because they are the only ones in a position to carry it.