Artificial intelligence is drastically reducing the time needed to analyse wildlife camera trap data, a process that traditionally takes conservation groups months or even up to a year.
A new study led by researchers from Washington State University (WSU) and Google, published in the Journal of Applied Ecology, demonstrates that AI can perform wildlife tracking analysis in just days, while producing scientific conclusions that closely match those derived by human experts.
Camera traps—motion-activated cameras placed in natural habitats—are widely used to monitor wildlife populations, generating vast amounts of image data. Projects can produce hundreds of thousands to millions of images that require painstaking review to identify species and their behaviors. Even with dedicated teams of assistants, this review process can delay analysis by six to seven months or longer, slowing conservation efforts.
The AI analysis matched human-derived models to about 85% accuracy
The study tested whether a fully automated AI system could replace human analysts in processing this enormous volume of data. The team used SpeciesNet, a general AI model developed by Google, to analyse images collected from diverse ecosystems including Washington state, Montana’s Glacier National Park, and Guatemala’s Maya Biosphere Reserve. They then compared the AI-generated results to traditional datasets labeled by human experts.
“We’re not trying to replace people,” said Daniel Thornton, lead author of the study and a wildlife ecologist at WSU. “The goal is to help researchers get to answers faster so they can make better decisions about managing and conserving wildlife.”
Dan Morris, a senior staff research scientist at Google and co-creator of SpeciesNet, emphasised, “The key question wasn’t whether the AI got every image right. It was whether the ecological conclusions you care about would end up being basically the same.”
The findings showed that for most species, the AI-produced models aligned with human-derived models in roughly 85–90% of cases. Key ecological measures—including species occupancy and the environmental factors influencing their presence—were consistent between AI and human analyses. Even when the AI made errors, such as misidentifying species or missing detections, the overall occupancy models remained robust because they rely on repeated observations over time, dampening the impact of individual mistakes.
Traditionally, early AI tools helped by filtering out blank images, which often account for 60–70% of camera trap data, but human experts still had to review tens of thousands of animal-containing photos. This study went further by eliminating the final human review step entirely, demonstrating that fully automated analysis is now feasible for many species.
Removing the analysis bottleneck for smaller research teams
The time savings are dramatic. What once took six to twelve months can now be accomplished in just a few days or about a week, cutting a major bottleneck in wildlife monitoring. This speed allows conservationists and wildlife managers to move more quickly from data collection to actionable decisions, potentially enabling near real-time monitoring of species such as jaguars, wolves, and grizzly bears.
This efficiency is particularly transformative for smaller or underfunded conservation groups, which often lack the resources to process large datasets quickly. Faster analysis also opens the door to expanding monitoring programs without being limited by processing capacity, enhancing the scope and scale of conservation efforts.
The research team also contributed to the wider AI-for-conservation community by making part of their dataset publicly available. Sharing data supports the refinement of AI tools like SpeciesNet, which improve through access to diverse and extensive training datasets.
Morris noted, “We weren’t trying to invent a new model. We were asking whether, given where the technology is today, people can rely on it for the kinds of analyses they already do.”
The AI analysis may be less effective for rarer species
However, limitations remain. The study focused on a subset of species commonly caught on camera. Rare or easily confused species still pose challenges for AI detection, and human review remains necessary for many other uses of camera trap data. Despite these caveats, the results suggest that image processing no longer needs to be a major constraint on large-scale camera-trapping studies.
“The big takeaway is that this doesn’t have to be a bottleneck anymore,” said Thornton. “If we can process data faster, we can respond faster, and that’s really what matters for conservation.”
Additional co-authors on the study include Travis King and Lucy Perera-Romero of Washington State University; Alissa Anderson of WSU and Montana Fish, Wildlife and Parks; Rony Garcia-Anleu from the Wildlife Conservation Society’s Guatemala Program; Scott Fitkin of the Washington Department of Fish and Wildlife; and Carly Vynne of RESOLVE, who all contributed to data collection, analysis, and manuscript development across the study sites in Washington, Montana, and Guatemala.
This study marks a significant step forward in harnessing AI technology to accelerate wildlife monitoring and conservation decision-making, promising a future where technology and ecology work hand-in-hand to protect biodiversity.