U.N. Experts Warn: The Perils and Promise of AI in Indigenous Conservation

through satellite imagery and sensors. “When integrated with Indigenous knowledge, AI can assist in predicting climate-related impacts, tracking wildlife movements, and enhancing land-use planning while improving resilience strategies,” she noted. In Brazil's Acre state, Indigenous agroforestry agents in the Katukina/Kaxinawá Indigenous Reserve have employed AI to address deforestation threats. This reserve is among the top five areas at risk for deforestation, as indicated by forecasts from an AI tool created by Microsoft in collaboration with the Brazilian NGO Imazon. Siã Shanenawa, one of the agroforestry agents in the reserve, communicated to Mongabay, “Monitoring our land is crucial because we Indigenous people are safer when we can spot if someone is encroaching, extracting wood, hunting on our territory, or igniting fires nearby.” Lars Ailo Bongo, a professor at the UiT The Arctic University in Norway and head of the Sámi AI Lab—a research group at the Sámi University of Applied Sciences and the Arctic University of Tromsø—expressed to Mongabay that while AI is still not inclusive enough, it offers some promising avenues for communities. He mentioned in an email that “AI can democratize access to the analytical tools necessary for conducting data-driven modeling that aligns with Sámi perspectives and values.” This advancement can facilitate more equitable arguments against organizations that possess the resources to create models that cater to their own objectives. In Nunavut, Inuit communities are merging traditional knowledge with predictive AI models and time-series analyses to identify new fishing areas affected by climate change. Likewise, in Chad, Indigenous pastoralists are utilizing participatory mapping alongside satellite data and predictive AI tools to foresee severe droughts and secure migratory routes, thereby enhancing their climate resilience. In South America, the Rainforest Foundation US is integrating traditional knowledge with modern technologies—from planting trees along boundaries to utilizing smartphones and drones—to assist Indigenous communities in safeguarding their lands. Cameron Ellis, the organization's field science director, stated to Mongabay via email, “AI is the latest addition to this continuum. Community monitors can utilize AI-generated remote sensing data to analyze extensive satellite information and interpret deforestation trends associated with mining or agricultural expansion, allowing for quicker responses to these challenges.” However, the deployment of AI also brings risks for Indigenous peoples. Residents and farmers in Thailand's Chonburi and neighboring Rayong provinces, plagued by water shortages and pollution, have expressed concerns regarding the environmental repercussions of expanding data centers—essential infrastructures that facilitate AI operations. These data centers demand significant amounts of water for cooling and substantial energy for functioning. Similar situations are happening across various communities worldwide, from rural areas in eastern Pennsylvania, USA, to villages in Querétaro, Mexico, where residents voice worries about wastewater contamination, water and energy shortages, and rising costs linked to data center growth. Ibrahim pointed out, “While AI is often viewed as intangible, it carries a very tangible environmental impact.” She explained that it relies on extensive energy, water, and critical minerals, many of which are sourced from regions near Indigenous territories, resulting in land degradation, biodiversity loss, and sometimes the displacement of communities. Beyond the environmental ramifications of data centers, Ibrahim's study also highlighted further challenges for Indigenous peoples linked to AI, such as insufficient infrastructure, legal safeguards, and institutional capacity to protect digital rights. She cautioned that AI might exclude Indigenous populations or enable the extraction of sensitive information. The deployment of drones, satellites, or mapping tools without prior consultation with Indigenous communities can jeopardize the confidentiality of sacred sites, ecologically significant areas, and other sensitive locations. Kate Finn, a citizen of the Osage Nation and executive director of the Tallgrass Institute—which aims to align investor strategies with Indigenous rights—identified what she termed “opportunity space” within AI to aid Indigenous peoples in preserving their languages and enhancing governance structures. However, she concurs with Ibrahim's concerns regarding environmental hazards, stating, “Indigenous peoples globally consistently demand that their free, prior, and informed consent be honored before data centers are established on their lands. There are numerous advantages and disadvantages to having a data center near one’s territory.” She further noted, “As we evaluate AI from an Indigenous perspective, it is imperative to consider all these various aspects, encompassing both opportunity spaces and the protective spaces of lands, territories, and resources, as well as language, culture, and the intellectual property that Indigenous peoples have shared online.” Bongo highlighted that the Sámi are constrained by a lack of funding to recruit AI developers capable of creating Sámi-oriented AI models and making them accessible to the community. He lamented, “This is particularly unfortunate, as there are Sámi AI developers eager to undertake this work,” indicating that the challenge is one of capacity rather than competency. “To advance, a larger initiative or support is necessary, which Sámi organizations currently lack the financial resources to achieve; thus, state funding from Norway, Finland, and Sweden is essential.” Additionally, for initiatives reliant on external funding, it is crucial to ensure that Indigenous peoples are not relegated to minority partners. Ellis emphasized, “Technology alone does not protect forests—people do.” He elaborated, “These tools are only effective when they are rooted in community governance and leadership, and when the data they produce is utilized to spur action locally. Furthermore, communities must maintain sovereignty over how their data is collected to ensure it serves their priorities without compromising their rights.” Ibrahim asserted that to safeguard Indigenous peoples and their territories, governments need to curb all forms of land theft, water exploitation, and mining operations tied to data centers and energy sources, while also respecting Indigenous rights, worldviews, and aspirations. “AI becomes detrimental when imposed without free, prior, and informed consent,” Ibrahim remarked. “In such scenarios, it risks perpetuating historical patterns of resource extraction, data appropriation, and the usurpation of knowledge and credit for this knowledge.