.png)
Here I am, frozen mid-doomscroll, staring at a cartoon strawberry in a plunging white shirt and teeny tiny miniskirt crouching to pick up a pen off of an office floor. A group of tall bananas in suits smirk and watch on. It has almost 900,000 views. Now, a sexy apple in a tight silk dress is swishing her way through a park — a strawberry man ogles her as his dowdy strawberry girlfriend berates him. It's been viewed over 200,000 times. Next, an avocado in tiny short shorts is brazenly ignoring her hardworking kiwi husband, before callously leaving him for a wealthy apple who showers her in diamonds. This last one has been viewed over 2 million times.
TikTok content
For the uninitiated few who have somehow, blissfully, managed to avoid any and all social media of late, this is the latest development in the never-ending churn of AI slop: the AI fruit drama. And people are eating it up — pun not intended.
Suddenly, our algorithms have filled with dramatic, fruity tales. It began with one-off stories starring a wide array of fruit characters: hot, young fruit ladies stealing other fruits' husbands, fickle money-hungry fruit women leaving their husbands, and frumpy fruit mums farting in bed. And now, in a new development, a series of “fruit Love Island” videos have sprung up, starring bikini-clad fruit ladies vying for a selection of hunky fruit bachelors. All of these videos are drawing in millions of views and increasingly unironic heated debates in the comment sections.
Glamour chats to the broadcaster about his most challenging subject yet: men.
The rise of the AI fruit drama is, at the very least, very annoying and, at the very worst, deeply concerning. Aside from the small matters of our brains seeping out of our ears and real artists losing more value and respect by the day, this particular brand of AI slop raises one very important question: why do AI fruit dramas hate women?
First of all, there's no denying that these videos are almost always drawing from deeply sexist stereotypes. They centre around “highly stereotyped narratives” that “often centre on a female-coded character — typically young and conventionally attractive — who is depicted as cheating on her partner with a male-coded character portrayed as wealthier or higher status,” says Alessia Tranchese, Associate Professor of Language, Feminism and Digital Media at the University of Portsmouth.
“What is striking is how closely these narratives mirror themes commonly found in manosphere discourse (especially in incel circles), particularly the idea that women exploit ‘good’ men for financial or social gain,” she adds.
Is AI designed this way? According to Tranchese, these videos are “largely produced by small or anonymous creators, many of whom rely on shared workflows and templates circulated on platforms like Reddit. These workflows are often reused and lightly modified, which is probably why many of the videos have a similar look and feel while still appearing slightly different from one another.”
But as to the question of whether these creators are actively seeking to make sexist content, it's less overtly sinister, but perhaps just as concerning. Of course, Love Island is filled with misogyny, she explains — because the real version is, too. “It's not surprising to me at all that we see some of those same misogynistic tropes playing out in the digital version,” she says. “These new technologies pick up on these existing sexist scripts in society, and they can amplify them.”
After all, AI learns from us — if it's generating more and more blatantly misogynistic stories filled with more and more two-dimensional sexist stereotypes, it's because it's learning that script from us.
And, of course, the way that viral AI slop works is by becoming more and more shocking — more and more extreme. After all, the more on the nose, the more we can't seem to look away — and the more we watch, the more it creates.
TikTok content
“It started off as something that people engage with quite ironically, and then became very normalised,” says Kerry McInerney, an AI ethicist and Lecturer at the University of Auckland. “The trouble is satirical engagement, humorous engagement, earnest engagement — it all comes down to clicks and engagement. And so then that does encourage the further production of AI slop. Even if people are engaging with it ironically, that could still potentially lead us towards a world where people are really incentivised to make more and more and more of this kind of content.”
“You don't have to be a millionaire influencer to access the manosphere; it's under our noses, all the time.”
“There was a narrative maybe several years ago that tools like AI could be used to somehow reduce or address gender equality — that they could make things like hiring less discriminatory and more fair, for example,” adds McInerney. “And I think we've just not seen that play out in practice. Instead, I think we've seen AI has this exacerbatory effect, and we see that in everything from, like, AI slop content like this, all the way through to kind of other really intentionally malicious uses of generative AI, like the production of pornographic deep fakes.”
Ultimately, the reason for the AI fruit drama's sexism comes down to a mixture of elements: “Creators [may be] drawing directly from online subcultures or deliberately reproducing such tropes because they are known to generate engagement,” says Tranchese. “It is also possible that the outputs are shaped by the training data behind generative AI tools, which may encode and reproduce existing biases, including content that aligns with what the models ‘predict’ will perform well with online audiences. All these dynamics might be at play at the same time.”
Ok, so AI fruit dramas are sexist — but does it really matter? Surely they're just dumb, meaningless slop videos we can write off as a brief blip in the ever-changing landscape of social media trends?
“They seem very innocuous,” says McInerney. “You know, they're not like some of the other AI-generated slop content that has been either very explicitly political or dangerous deepfake content used to harass and abuse. This feels small fry, comparatively; it seems really innocuous.”
But that doesn't mean we should normalise — or watch, even ironically, “The normalisation of these outdated gender stereotypes is being given new life in AI slop. It's easy to mindlessly consume it and share it and not think very hard about it. But then that does contribute to a culture where this kind of behaviour and these ideas are normal.”
Today – and every day – the women fighting for a safer future deserve their flowers.
Adds Tranchese, “It's important to see these videos as part of a wider content ecosystem rather than as an isolated trend. Today it may be AI-generated fruit videos; tomorrow it will be something else. However, many of these trends recycle similar underlying narratives, often portraying women as inconsistent, superficial or emotionally manipulative. While individually trivial, these posts repeatedly reinforce the same sexist assumptions.”
She adds, “AI itself is not inherently harmful, but when it is deployed within broader societal and economic contexts that already contain inequality — in this case a patriarchal society that exploits inequality for profit — it can end up reinforcing those patterns.”
AI was meant to help free us from mundane tasks and open up new space and time for creativity and deep thought. Instead, I'm sitting here watching a strawberry get slut-shamed. Forgive me if I sound as bitter as a lemon, but this is seriously depressing and we deserve better.
