A house cat stares down a leopard in a suburban backyard. The big cat growls, the child in the background freezes, and—against all odds—the cat lunges. The leopard turns tail and bolts. Within seconds, the clip racks up thousands of likes and comments: “Cats really are fearless!” or “This is why I love animals!”
Except, of course, none of it ever happened.
The leopard, the child, and the heroic cat are all creations of artificial intelligence. Yet the video has been shared tens of thousands of times across TikTok, Facebook, and X, fooling millions into believing it’s real. And it’s not alone. Other viral AI-generated wildlife clips show raccoons riding crocodiles down jungle rivers, bears playing on trampolines, and deer politely joining family picnics. They’re funny, heartwarming, and almost believable—but according to scientists, they’re also dangerous.
A new study from the University of Córdoba (UCO) in Spain warns that this wave of synthetic wildlife videos is quietly eroding public understanding of nature. Far from being harmless entertainment, these ultra-realistic fabrications are distorting how people—especially children—think animals behave, what species are truly endangered, and even what kinds of creatures live near them.
“The main conclusion from our study,” José Guerrero Casado, lead author of the research, explains to A-Z Animals, “is that AI-generated wildlife videos often present distorted or unrealistic portrayals of nature. These misrepresentations can spread misinformation, especially among children and audiences with limited ecological knowledge.”
As social media becomes a dominant source of environmental learning, the researchers say this flood of digital fantasy is undermining conservation efforts and fueling a dangerous disconnect between society and the wild world it claims to love.
Why Are AI-Generated Wildlife Videos Harmful?
How wildlife is communicated about and seen across social media has large implications for conservation.
©Julia Ardaran/Shutterstock.com
To understand the implications AI-generated wildlife videos have on audiences, the Córdoba team analyzed dozens of viral AI-generated clips across the social media platforms TikTok, Facebook, Instagram, and X, examining how often they were shared, liked, or commented on to measure their influence. They found that these digital creations often depict animals acting in ways that defy biology and ecology, predators playing gently with prey, apex carnivores wandering suburban lawns, and rare species appearing as common backyard visitors.
“Some AI-generated videos depict highly implausible or exaggerated human-wildlife interactions—such as large carnivores entering homes, attacking pets, or even harming people,” Casado says. “These videos are often presented in a format that imitates low-quality security footage, which can make them appear more authentic.”
Such fantasy-filled depictions can have subtle but serious consequences. When young audiences or less scientifically literate viewers take these portrayals at face value, they form unrealistic expectations about the natural world. “When social media users, especially children or individuals with limited ecological knowledge, perceive these videos as real, they may develop distorted views of wildlife,” adds Casado.
The problem lies not only in spreading falsehoods but also in creating a false sense of familiarity. Videos that portray dangerous species as friendly or endangered species as abundant can distort the sense of rarity and risk that conservation messaging depends on. They erode the emotional urgency that motivates people to protect wildlife.
The viral “cat versus leopard” clip is a particularly damaging example. The idea that a leopard would appear in a European suburb—or that a domestic cat could scare one away—is pure fiction, but the imagery feels plausible enough to shape viewers’ beliefs about big cats.
This distortion becomes more dangerous as AI tools grow more accessible. Videos that once required film studio budgets can now be created on a smartphone in seconds.
“Some AI-generated videos depict highly implausible or exaggerated human-wildlife interactions—such as large carnivores entering homes, attacking pets, or even harming people,” Casado explains. “These videos are often presented in a format that imitates low-quality security footage, which can make them appear more authentic.”
The psychological effect may swing in two directions. On one hand, fake videos showing predators in human spaces can heighten fear, making species like leopards, wolves, or sharks seem more threatening than they are. That, in turn, could reduce public sympathy for conservation measures. On the other hand, overly cute or humanized portrayals of wildlife—animals acting like pets or performers—could make viewers complacent, leading them to believe that endangered species are plentiful or that wild animals are safe to approach. Both extremes weaken conservation attitudes and blur the line between wildness and domestication.
These videos are often presented in a format that imitates low-quality security footage, which can make them appear more authentic.
José Guerrero Casado, professor of zoology at University of Córdoba
A New Kind of Conservation Challenge
Spotting fake videos is key to helping stop the spread of misinformation.
©Andrey_Popov/Shutterstock.com
Casado sees this as a new frontier in conservation challenges: misinformation masquerading as nature. The danger, he suggests, lies not just in the loss of ecological understanding but in a broader societal disconnection from the real natural world.
“Previous research, including our own, highlights a concerning disconnect between society and the natural world—particularly in terms of knowledge about native wildlife. For example, children often recognize exotic charismatic mammals more easily than common native species found in their own surroundings,” he says.
He believes education must be at the center of the solution. “Education plays a fundamental role in addressing this disconnect,” Casado emphasizes. “By integrating more content about local ecosystems and native fauna into curricula, we can foster a deeper understanding and appreciation of the biodiversity that surrounds us. This foundational knowledge is a crucial previous step to ensure positive attitudes toward conservation.”
Media literacy is the other half of the remedy. Casado argues that just as people are taught to identify fake news, they should also learn to question synthetic nature media. “When users encounter suspicious wildlife videos,” he says, “they should be encouraged to verify the information using reliable sources—such as scientific publications, environmental agency websites, reputable NGOs, or even trusted social media profiles that specialize in nature education.”
He also calls on tech platforms to take greater responsibility. While users can manually label posts as AI-generated, Casado believes automated systems should do the same. “Social media should develop a tool to automatically detect a video or picture created by AI, and then label this post as AI-created,” he says. “This would help prevent the spread of misinformation and reduce the likelihood of misleading portrayals influencing public attitudes toward wildlife.”
Next Steps: Looking at Emotional Reactions
Guerroro Casado and his colleagues are hoping to dive further into this issue.
©PeopleImages/iStock via Getty Images
Looking ahead, Casado and his colleagues plan to expand their analysis by studying how people react emotionally to AI-generated wildlife content. “We would like to test if people are able to distinguish real videos and those created by AI, and what they feel when they see the content,” he explains. By examining responses to fabricated videos—such as clips of humans ‘rescuing’ wild animals or keeping them as pets—the researchers hope to better understand the psychological and cultural impact of synthetic wildlife imagery.
For Casado, the issue ultimately comes down to truth in representation. Wild animals, he says, deserve to be seen for what they truly are—not as digital caricatures, internet memes, or fantasy companions. “AI-generated media tends to anthropomorphize wild animals—portraying them as if they behaved like humans or could be domesticated,” he notes. “It’s important to emphasize that wild animals are not pets. Their beauty lies in their wildness—in their natural behaviors and interactions within ecosystems.”
As artificial intelligence continues to reshape how humans experience the natural world, Casado offers a simple but urgent reminder: “A well-informed society, one that understands wildlife through accurate, science-based information, is essential to support conservation actions.”
In an era when pixels can pass for reality, protecting wildlife may now depend as much on defending truth as on defending the animals themselves.
