A cartoon fruit series generated entirely by artificial intelligence has become one of TikTok's most-watched properties of 2026, prompting researchers and critics to warn that platforms are systematically rewarding low-quality, harmful content at scale.

Fruit Love Island—an AI-created spoof of the reality TV show with human-like animals, such as Strawberina and Bananito among other characters, has been viewed 300 million times in nine days since its premiere in March, which most of its viewers claimed that they have discovered unintentionally. TikTok’s recommendation system suggested the show to many users even before it became clear that the videos were being quietly removed from the website, keeping only 10 out of 22 original episodes online. 

Critics have flagged the content weeks earlier. Commentators described episodes as "littered with sexist stereotypes", with female-coded characters routinely humiliated or placed in physical altercations played for laughs—content that many said they found uncomfortable but kept watching regardless. TikTok did not respond to requests for comment on why the content remained algorithmically promoted throughout.

Researchers say the series' virality is a symptom of a structural problem, not an anomaly. "Shows like Love Island rely on simple archetypes and predictable dynamics, which makes them especially easy for AI to replicate and remix at scale," said Michael Grabowski, professor of communication at Manhattan University. The result is content that is cheap to produce, easy to consume and hard to look away from—even when viewers aren't sure why.

The phenomenon has a name in online spaces: “AI Slop”. It refers to generative AI content produced at volume with little creative intent, optimised for engagement rather than quality. According to Know Your Meme, AI fruit videos have been circulating since February 2026, when an account began posting videos about a strawberry's extramarital affair with her eggplant boss. Fruit Love Island industrialised the format and an algorithm did the rest.

The series' creator cited harassment and lost revenue when announcing its end on 28 March, posting expletive-laden videos directed at critics before going silent. 

What remains is a question the series' 300 million viewers probably weren't asking while they were watching: if a single anonymous account can generate that kind of reach in under two weeks using AI, and platforms only act after sustained public pressure, are content moderation frameworks governing short-form video structurally unequipped for what comes next?

Image source: Wired