Founder’s Briefs: An occasional series where Mongabay founder Rhett Ayers Butler shares analysis, perspectives and story summaries.
Conservation journalists are facing a new issue: AI-generated wildlife imagery.
The issue is not just that fake images exist. That has long been true. What has changed is how convincing synthetic wildlife photos and videos have become, how cheaply they can be made, and how quickly they can spread. A clip can move through Facebook, WhatsApp, TikTok, or even LinkedIn before anyone has checked whether it shows a real animal, a real place, or a real event.
That matters because wildlife images carry an implicit claim. A photograph of a rare animal, a camera-trap still, or a video of unusual behavior usually tells the viewer: this happened. As generative AI improves, that assumption needs more scrutiny.
The risks are not theoretical. False videos of animal attacks can deepen fear in places where human-wildlife conflict is already difficult to manage. Fabricated images of wild animals behaving like pets can feed demand for the exotic pet trade. Misleading footage of rare species can absorb the time of researchers, journalists, NGOs, and public agencies that have to determine whether an event actually occurred.
It also changes the work of newsrooms. At Mongabay, we now spend more time looking at sourcing, provenance, metadata, reverse-image searches, forensic tools, and whether a photographer, researcher, or institution is known and trusted. AI detectors can occasionally help in some cases, but they cannot settle the question. False positives and false negatives remain common.
AI is already useful in conservation. Researchers use it to analyze camera-trap images, satellite data, and bioacoustic recordings. The problem begins when AI is used to fabricate an event and present it as observed reality.
For journalism and conservation, trust is part of the work itself. People need to know what species exist, how they behave, and what threats they face. As synthetic media becomes more common, careful sourcing and editorial judgment will matter more. So will the burden of showing that something extraordinary is real.
Rhett Ayers Butler recently spoke with Hannah Kaplan for the Wildlife Conservation Society’s Wild Audio podcast about AI-generated wildlife imagery. Listen to the podcast here.
Banner image: AI generated image found on Facebook. Label added by Mongabay.
Credits
