AI-Generated Far-Right Memes Are Flooding Facebook. Here Is How the Machine Works.

Key points

• AI tools now let bad actors produce hundreds of politically charged memes per day with almost no human effort.
• These posts spread most effectively on Facebook, where algorithmic recommendations, the automated systems that decide what you see next, favour content that triggers strong reactions.
• The content follows a recognisable formula: outrage headline, AI-generated image, minimal factual grounding.
• Researchers and journalists have documented a rise in what they call "AI slop", low-quality AI-generated content churned out purely for clicks and engagement.

First Dog on the Moon, the long-running satirical cartoon strip published by The Guardian, recently turned its lens on a trend that is anything but funny: meme factories using artificial intelligence to flood Facebook with far-right political content.

The cartoon is pointed satire. The underlying story is real.

What exactly is an "AI slop" meme factory?

Think of it as a content assembly line, except humans barely touch it. Someone sets up an AI image generator, the kind of tool that produces a finished picture from a short text description, and pairs it with an AI writing tool. The system then pumps out dozens or hundreds of posts a day, each one designed to look like a grassroots political opinion.

The posts often carry a recognisable shape. An alarming headline. A generated image of a crowd, a flag, or a political figure. A short caption built to make you angry enough to share.

None of it requires a journalist, a photographer, or even much time.

Why does Facebook keep showing this content to people?

Facebook's recommendation algorithm, the software that chooses which posts appear in your feed, rewards engagement. Likes, shares, and angry reactions all count as engagement. Content designed to provoke strong feelings gets pushed to more people, which generates more reactions, which pushes it further still.

That feedback loop is not new. What has changed is the supply side. AI tools have slashed the cost of producing emotionally charged content to nearly zero.

What should readers watch for?

A few signals tend to appear together in AI-slop posts. Images that look almost photographic but have odd details: strange hands, blurry background text, or faces that seem slightly off. Captions that use sweeping emotional language but cite no source. Account profiles that post many times a day, every day, often on the same theme.

If a post makes you want to share it immediately, that is worth a pause. That reaction is usually the point.

Suspicion is not paranoia here. It is just good reading hygiene in an environment where the volume of generated content keeps rising and the tools to create it keep getting cheaper and easier to use.

Common questions

Is this content illegal?

In most countries, generating or sharing political memes, even misleading ones, is not automatically illegal. Where it crosses into illegality is when it involves impersonating real people, election interference, or coordinated inauthentic behaviour that violates platform rules.

Can Facebook detect and remove this content automatically?

Facebook uses automated systems to catch some AI-generated content, but the tools that create the posts are improving at roughly the same pace as the tools that detect them. Human review teams remain part of the process, though critics argue the scale of the problem outpaces current moderation capacity.