Beyond the Hype: Navigating the Conservation Implications of Artificial Intelligence

Abstract

Conservation AI—the deliberate application of artificial intelligence technology to achieve conservation goals—has great potential to boost productivity, make existing conservation actions more efficient, and enable entirely new areas of activity. However, it also comes with risks, including AI being used by bad actors; high material demand for energy, land, and water; biases in training datasets; AI-fueled techno-optimism distracting from other actions; and undesirable changes in staffing and working practices in the conservation sector. Changes in wider society brought about by AI in areas such as agriculture, human health, and labor markets may also have significant impacts on biodiversity (whether positive or negative), as these are major drivers of biodiversity loss. This article reviews the various links between AI and conservation, arguing that to date there has been too much techno-optimism and a lack of attention to risks and broader implications. It concludes with recommendations for how conservation could approach AI more effectively by considering risks and potential unintended consequences; adopting a principle of transparency; ensuring AI does not harm the staff, skills, and independence of the conservation sector; and investing in research and advocacy to address the conservation implications of wider societal changes caused by AI.