Chadian activist Hindou Oumarou Ibrahim on Indigenous rights in AI governance

• Hindou Oumarou Ibrahim argues that AI governance must treat Indigenous peoples as rights holders and decision-makers, not merely as stakeholders.
• Current AI policies often prioritize Western-centric risks like misinformation, neglecting the impacts of AI infrastructure on biodiversity, water, and land.
• Indigenous communities advocate for the protection of collectively owned data and environmental rights across the entire AI value chain, from mineral extraction to tool deployment.

Concerns with Current AI Models

• AI data centers consume significant energy, water, and land, often encroaching on Indigenous territories.
• Hardware production relies on mineral supply chains that can drive ecosystem destruction.
• Traditional data protection frameworks focus on individual privacy, failing to account for the collective nature of Indigenous cultural, linguistic, and environmental knowledge.

Recommendations for Governance

• Governments should require social and environmental impact assessments for all AI infrastructure projects.
• Policies must uphold the principle of free, prior, and informed consent regarding any technology affecting Indigenous lands or intellectual heritage.
• AI companies should audit models for linguistic and cultural bias and include Indigenous experts in the development process.

AI as a Tool for Conservation

• When directed by community priorities, AI can strengthen rather than replace traditional knowledge.
• Successful examples include:
  • In Chad, participatory mapping and predictive AI help anticipate drought and protect herders' migration routes.
  • In the Amazon, AI-supported satellite monitoring detects illegal mining and deforestation.
  • In the Arctic, Sámi and Inuit communities use predictive models to monitor wildlife and changing sea ice conditions.