For a while, AI slop was mostly harmless filler: fake animals, invented recipes, videos made to farm clicks. It has now found politics. A wave of AI-generated “news” channels on YouTube is manufacturing stories of economic collapse in Democratic-run states. Some are pulling hundreds of thousands of views.

The videos follow a formula. They claim big companies are fleeing California. Shops are running out of food. Teachers are quitting in droves. One channel, @MeganWright62, posted a false claim that Walmart was closing hundreds of California stores. CNN debunked it. In January, President Trump shared it anyway.

Much of the wave targets Gavin Newsom, a likely 2028 candidate. His team told Semafor it has worked with YouTube to remove more than 700 accounts since January. It used the platform’s “likeness protection” system, which flags a face used without permission. More than 100 carried false claims about Newsom and California.

The scale is the point. One channel, New York Reports, ran a video headlined “Wells Fargo LEAVES to Florida.” Stitched from unrelated clips, it drew about 300,000 views. Others falsely claimed Target had shut all its New York shops and that JP Morgan had quit the city. After Semafor flagged them, YouTube pulled several for breaking its spam rules.

The platform’s problem

YouTube says it is trying. Its chief executive, Neal Mohan, has called fighting AI slop a “top priority”. The company removed 1.6 million accounts for spam last year, and is tightening the rules on monetising repetitive, template-made videos. But the flood keeps rising. One study found more than 20% of the videos YouTube shows new users are low-quality AI content.

The trouble is the incentive. Slop exists to harvest cheap views and the ad money that follows. Pulling one channel does little when another appears the next day. It is the same dynamic seen elsewhere. LinkedIn added a “slop” button. Snapchat stopped recommending AI clips. Both are part of a wider retreat from features the industry rushed out.

This particular wave hits Democrats, but the picture is not one-sided. YouTube says it enforces its rules regardless of politics. Republicans have been quicker to use AI in campaign messaging. Democrats have faced backlash for it, as Andrew Cuomo did over an AI ad. And conservative media has made human versions of “blue-state dystopia” content for years.

Why it matters

Before the last US election, the fear was the perfect deepfake, a fake video convincing enough to swing a vote. That threat was largely overblown. The real problem turned out to be volume. Not one flawless fake, but an endless tide of cheap, plausible-enough clips. They are hard to trace and easy to believe at a glance.

Researchers argue the fix is not a blanket ban on AI content. Small tweaks to ranking, monetisation and eligibility could shift what creators make. New rules in Europe now force platforms to label synthetic media. For now, though, the slop still pays. And the people it targets are the ones hoping to run in 2028.

Get the TNW newsletter

Get the most important tech news in your inbox each week.