From Talking Bananas to Digital Gore: How AI Tools Fuel Viral Violence

Reading GNET's latest brief on a Tuesday morning, I kept circling the same image: the same free apps letting teenagers animate talking bananas into soap-opera betrayals are now producing something uglier — cartoon fruit, diced and tortured in high-definition. According to a new insight from the Global Network on Extremism and Technology, that surreal corner of TikTok has quietly evolved into what researchers call "AI fruit/vegetable gore," a subgenre where colourful, anthropomorphic produce characters are subjected to extreme mutilation and hyper-realistic violence, all generated by tools anyone can sign up for in under a minute.

The economics behind the slop

The report doesn't dwell on aesthetics; it goes straight for the business model. Tools like VeoE.ai, PixVerse, Sora, Media.io and Grok all offer free registration and a limited free tier — a few images, a short clip, maybe a trial week — before paywalls kick in. That frictionless onboarding is the engine. A single creator can pump out hundreds of short videos cheaply, flood platforms with them, and let the algorithm sort winners from losers. GNET calls the result "AI Slop": low-quality, high-volume synthetic content engineered for clicks, watch time, and ad revenue.

This is where the "brain rot" framing earns its keep. Brain rot isn't the content itself; it's the cultural exhaust of consuming absurd, fragmented, overstimulating media in an infinite scroll. Anthropomorphic fruit locked in melodramatic love triangles, betrayal, and domestic warfare is the perfect delivery mechanism — harmless-looking, meme-shaped, and algorithmically magnetic.

When the algorithm learns to weaponise cute

Then the report turns darker. Researchers monitoring the trend identified an increasingly violent offshoot. In these videos, the same cartoon strawberries, lemons, and carrots that viewers laughed off a week ago are now subjected to extreme mutilation and torture, rendered with the same accessible tools — Viggle.ai is named specifically — and distributed across TikTok despite the platform's prohibitions on violent content.

The pattern matters because it tracks how radicalisation now travels: not through manifestos, but through formats that look absurd, ironic, and shareable. Repetitive, hyper-stimulating storytelling normalises aggression by making it familiar, funny, and endlessly remixable.

Where the business model starts to bleed

For platform teams, the uncomfortable takeaway is that volume moderation isn't the same as content moderation. The economics reward flooding; the algorithm rewards engagement; violence merely rides the wave. The infrastructure questions here overlap with work on how blockchain technology is reshaping digital entertainment ecosystems — attempts to attach verifiable fingerprints to synthetic media at the moment of creation, before the meme factory spins it into something else.

Whether any provenance layer survives contact with a thousand-person content farm is the open question worth tracking. For now, the fruit is still cute. The knives are just getting sharper.