Artificial intelligence is emerging as both an opportunity and a risk for African biodiversity conservation, with scientists warning that uncritical use of AI tools could undermine ecological insight and reinforce existing biases — even as the technology promises new ways to process the overwhelming volumes of data the field relies on.
Conservationists routinely analyze decades of weather data, the movements of millions of insects and other ecological information. Until now, they have largely had to find and sort information manually before applying statistical tools that often oversimplify the source material. AI tools now promise to help — but they are far from perfect. They can confidently fabricate information and amplify hidden biases in their training data, and different AI tools have different uses, strengths and weaknesses.
AI featured among the top 10 emerging issues in biodiversity conservation in South Africa in a recent horizon scan undertaken by a group of 14 experts in the field. Drawing on discussions across their professional networks, literature and news trends, the panel identified issues likely to emerge and intensify over the next 5 to 10 years. The issues fell into three main groups — technological disruption, regulatory complexity and infrastructure impacts — with AI cutting across all three.
The scan highlighted significant opportunities. One potential use of AI is tracking. Tracking animals and insects at scale is essential for conservation decisions — birds and whales migrate across the planet every year, and insect populations shift through the seasons in the billions. Image recognition AI can process camera trap data to populate databases such as Wildlife Insights and produce information about animal behaviour to help predict the impacts of climate change and industrial development on biodiversity. Mass monitoring also captures people sharing those landscapes with animals, and the surveillance can be used to detect illegal wildlife harvesting or reduce human-animal conflict.
Land use is another area where AI offers opportunities. Using economic data together with landscape information, custom AI models can be trained to predict deforestation, allowing preventive action, or to identify land with high conservation value at the best available price. AI can also help condense ecosystem complexity into maps and categories to inform broad landscape-level decisions, increasing the volume of data that can be summarized.
Chatbots in particular can distil information from huge amounts of text. They can monitor product listings to detect illegal wildlife trade online the moment it occurs, read hundreds of scientific publications to help decide which species are at risk of extinction, and draw on many sources to draft environmental impact assessments — offering a tempting shortcut around a time-consuming reporting task.
But the scan also identified significant risks. Local communities living off the land might experience mass surveillance as an intrusion. Alienation of local communities in that way could cause them to oppose conservation governance and even sabotage technology in the field to protect their privacy.
The technology itself has limitations too. Using AI for tracking animals means specially training image and audio identification systems to work with each ecosystem and piece of hardware. An AI model is only as good as the effort put into training it — feeding it city recordings, for example, may cause it to “hear” pigeons everywhere and produce confident but incomplete species lists from natural-area data. Another concern is that replacing human involvement could contribute to job losses and to an ongoing decline in taxonomy knowledge, a decline that is more severe in biodiversity-rich, low-income countries in Africa. That knowledge is essential for improving and correcting AI systems in the first place.
The land use applications raise similar concerns. AI tools used for mapmaking risk disconnecting maps from ground reality by replacing human judgement in the field and favouring data sources compatible with AI methods. A skilled ecologist surveying an ecosystem will notice unexpected details that were not specified at the planning stage — speaking with local people, for example, may reveal planned farming expansion or local harvesting practices. An AI system would miss that context because it can only read information that has been digitized. It also cannot see animals that evade cameras or identify species it was not trained on, and it cannot speak to humans to discover their intentions or uncover ecological wisdom passed down through generations.
Chatbots present their own risks. They can generate or embed fictional information, and even when drawing on real material they often reflect bias in their training data — favouring research and perspectives from well-represented institutions in the Global North, where publications have historically been dominated by men in high-income universities. Uncritical use of chatbot-generated recommendations could lead to poor environmental decisions, such as suggestions to plant trees without considering diverse ecosystems like Africa’s savannah grasslands. Using chatbots as a shortcut to summarize knowledge and inform conservation decisions in Africa risks reinforcing colonial systems and marginalising indigenous communities and knowledge.
Strong regulation of AI in environmental science is therefore both a moral and legal imperative, the authors argue. The sector needs clear safeguards, standards and oversight mechanisms to prevent faulty or inappropriate AI outputs from influencing decisions — including validation protocols to catch fabricated information, limitations to prevent chatbots from overriding human knowledge and perspectives, mandatory disclosure of AI prompt histories, and standards for describing training datasets so that appropriate models can be selected.
The explosion of AI presents a powerful opportunity for conservation if used carefully. If unchecked automation replaces human judgement, the authors warn, conservation risks becoming the tool of the very systems it has been built to challenge.
The analysis was authored by Jeran Cloete, a PhD candidate in conservation ecology and entomology at Stellenbosch University; Dian Spear, senior research scientist at Stellenbosch University; Jessica da Silva, principal scientist; Lavhelesani Dembe Simba, lecturer in entomology at the University of Fort Hare; and Peter J Carrick, honorary research fellow at the Institute for Communities and Wildlife in Africa at the University of Cape Town and founder and director of Nurture Restore Innovate.
