Key facts
- AI-generated animal videos are widely shared online, accumulating millions of views.
- Researchers from the University of Cordoba published a paper in Conservation Biology on the issue.
- Fabricated footage often depicts implausible interactions between species or exaggerated human-animal bonds.
- AI content may disproportionately focus on mammals, potentially diverting conservation funds from other groups.
- AI-manipulated images submitted to citizen-science platforms can contaminate scientific data.
- A separate study in Nature Ecology and Evolution highlights the risk to citizen-science data.
Researchers are warning that the proliferation of AI-generated animal videos online is creating a distorted perception of wildlife behavior and could be detrimental to conservation efforts. A study published in Conservation Biology by researchers at the University of Cordoba highlights how convincing fabricated footage is increasingly shaping public understanding of animals.
The paper points to examples such as AI videos showing birds sheltering chicks, which inaccurately suggest parental care is solely maternal, ignoring the significant role males play in many species. Fabricated interactions between predators and prey, or between humans and wild animals, are also cited as examples of misleading content. These scenes risk fostering a false sense of security around wild animals and could inadvertently fuel the demand for exotic pets, thereby contributing to the illegal wildlife trade.
Furthermore, the researchers anticipate that AI-generated content will likely favor mammals due to their popularity on social media, potentially exacerbating existing funding imbalances in conservation, where less charismatic species already struggle to attract support. The study also notes the risk of fake, location-tagged wildlife footage misleading tourists and placing undue pressure on sensitive ecosystems.
Echoing these concerns, a separate warning from researchers in Nature Ecology and Evolution suggests that AI-manipulated media submitted to citizen-science platforms could contaminate valuable ecological data. While outright fabrication is a concern, over-enhancement of images using AI tools can also alter identifying features, leading to misidentification of species. The researchers emphasize the need for greater media literacy and awareness among content creators and audiences to mitigate these risks, as global regulation of AI content remains unlikely in the near term.