Conservation International is deploying an integrated network of camera traps and acoustic sensors alongside artificial intelligence to monitor biodiversity in Cambodia’s Cardamom Mountains, transforming how researchers track endangered species in one of Southeast Asia’s most threatened rainforests. The initiative addresses a critical conservation need in a region that has lost nearly 7,000 hectares of tree cover over the past five years, primarily due to infrastructure development and historical deforestation pressures. The technology ecosystem relies on strategically placed passive monitoring devices supplemented by machine learning. Camera traps were systematically deployed across the Central Cardamom region, yielding a 2024 survey that documented over 100 resident species, including nearly two dozen classified as vulnerable or endangered. To track canopy-dwelling species that evade visual surveillance, researchers introduced bioacoustic monitors positioned at three-kilometer intervals. These devices capture vocalizations from gibbon troops without overlapping territories, generating extensive audio datasets. Data processing is accelerated through custom artificial intelligence models. Conservation International trained a machine-learning algorithm using approximately half of the collected acoustic data, teaching the system to distinguish gibbon calls from ambient forest noise. The AI subsequently automated the classification of the remaining 800 recordings in just six weeks. Project leadership plans to expand the model’s capabilities to differentiate vocalizations by sex and eventually identify individual animals, enabling long-term population tracking without invasive fieldwork. Ratha Sor, biodiversity and science manager at Conservation International, emphasized that the integration of hardware monitoring and algorithmic analysis provides verifiable proof of ecosystem health. The presence of sensitive species such as pileated gibbons, pangolins, and elephants signals that the protected area retains sufficient ecological integrity to support complex food webs. Indigenous community members, including local guide Pan Sok from the Chong minority, play a pivotal role in device deployment and maintenance, combining traditional ecological knowledge with modern survey methodology. While poaching and small-scale encroachment have declined due to enhanced ranger patrols, large-scale infrastructure projects, particularly hydropower dams, continue to drive habitat fragmentation. Conservationists aim to leverage the documented biodiversity data to advocate for strict environmental safeguards and secure long-term funding for the 1 million-hectare protected landscape. The expanded AI monitoring framework is scheduled for field validation later this year, with researchers preparing to refine algorithmic accuracy and scale deployment across additional transects. The project demonstrates how automated acoustic analytics and computer vision can provide scalable, low-impact solutions for monitoring remote ecosystems amid accelerating climate and developmental pressures.