AI-generated wildlife videos generate confusion and threaten conservation efforts

A study by the UCO advocates for media literacy and training for students to address the problems arising from the dissemination of AI-generated biodiversity videos

A video is circulating on social media in which a leopard enters the backyard of a house where a child is playing, and a cat confronts the leopard and scares it away, protecting the child. The video has garnered over a million “likes” and has been shared more than 15,000 times. Another series of videos shows bears or deer jumping on a trampoline in a back yard. In another, three raccoons float down a river riding on three crocodiles. All of these videos are created by artificial intelligence and, thanks to their high degrees of realism, appear to be real.

Given the proliferation of these types of videos and the lack of research on this topic in scientific literature, researchers from the GESBIO group at the University of Cordoba—José Guerrero, Francisco Sánchez, Antonio Carpio, Rocío Serrano, and Tamara Murillo—have spotlighted this issue, analyzing the different consequences that these AI-generated videos have on the knowledge and conservation of wild species. The main problems they have detected after analyzing the most shared videos on social media are: misperceptions of wild animals (it is highly unlikely that a leopard would jump into a backyard, and that a cat would confront it), the attribution of human characteristics and behaviors to animals, and the increasing disconnect between society and the natural world.

“They reflect characteristics, behaviors, habitats, or relationships between species that are not real. For example, we see predators and prey playing. They show us animals with human behaviors that are far from reality,” explains José Guerrero. “The video of the child playing in the yard, the one featuring the leopard, undermines the conservation of a species like this, as you will never encounter it in that situation,” he continues.

“There is already a total disconnect between citizens and wildlife, which is particularly pronounced among primary school children, as we saw in the IncluScienceMe project, which demonstrates a lack of knowledge of local fauna among young children. These videos create false connections with nature, as vulnerable species appear more abundant in these videos, and that is negative for conservation,” points out Rocío Serrano. These videos also cause frustration among young children, who expect to go out into the countryside and find a capybara, or have a wild animal come up to greet them. “If young children go out into the countryside and don't find these animals with more charismatic or magical characteristics and behaviors, it has the opposite effect in terms of connection,” she emphasizes.

Considering that during childhood the main form of learning is through images, this creates a distorted image. In addition, the increased use of social media as a source of information (especially among young people) exacerbates the negative impact of these videos. 

“We are faced with another serious problem: the demand for exotic species as pets,” says Tamara Murillo. Exposure to these exotic and charismatic animals, which are assigned sociable temperaments, makes more and more people want to have them at home.

The strategies proposed by the team include media literacy providing citizens with tools to question and verify information by turning to reliable sources and “trying to introduce environmental knowledge into school curricula, clarifying concepts such as native, what an exotic species is, and ensuring that children understand from an early age that there are no lions here,” recalls Francisco Sánchez.

This qualitative work, published in the Conservation Issues section of the journal Conservation Biology, opens up a line of study that had not been explored until now and promotes further research into the effects of AI-generated content on biodiversity conservation.

Reference:

Guerrero-Casado J, Murillo-Jiménez T, Carpio AJ, Tortosa FS, Serrano-Rodríguez R. Threats to conservation from artificial-intelligence-generated wildlife images and videos. Conserv Biol. 2025 Sep 3:e70138. doi: 10.1111/cobi.70138