Two Faces Of Same Algorithm: Algorithmic Disparate Impact On India's Adivasi Communities

India's forest governance is increasingly caught in a profound technological contradiction. While tribal welfare agencies deploy digital mapping tools to secure land tenure for marginalized forest dwellers, state forest departments deploy automated AI surveillance systems that criminalize the traditional, legally protected forest-dependent livelihoods of those exact same populations. This structural friction—where identical technological capabilities deployed by different branches of the same state yield diametrically opposed legal and social consequences—creates what is known as algorithmic disparate impact. The failure of coordinated governance leaves public-sector artificial intelligence (AI) open to being employed as a device of dispossession through conservation. These uncoordinated interventions are not meant to rectify the imbalance of the past but instead pose the possibility of repeating the paradigm of 'fortress conservation' that the Indian Parliament attempted to overcome through forest rights laws.

Institutional Contradictions: Two Ministries, One Forest

At the center of this algorithmic friction lie the conflicting administrative mandates of two primary public institutions. On the legal recognition side, the Ministry of Tribal Affairs (MoTA) has moved to streamline the arduous claims-verification process under the Scheduled Tribes and Other Traditional Forest Dwellers (Recognition of Forest Rights) Act, 2006 (commonly known as the Forest Rights Act or FRA). Rather than relying on static questionnaires like the older GIS-Enabled Entitlement Tracking System (GEET), MoTA has worked to build an AI-powered FRA Atlas and WebGIS-based Decision Support System. This sophisticated spatial database is an evolution from the Smart India Hackathon 2025, which was further enhanced by gathering information from the field in 2026. The introduction of this platform became a part of the revolutionary Tripura Digital FRA Atlas project. By employing GIS mapping in combination with field-verification data, this platform helps the Adivasis to generate spatial maps of their lands and thereby establish their ownership of these lands.

Contrarily, the forest departments at the state level, which function within the ambit of the Ministry of Environment, Forest and Climate Change (MoEFCC), use surveillance technology for a completely new purpose: monitoring forest violations and conservation boundaries. A notable example is GAJ-DASTAK, an AI-powered elephant detection and deterrence system developed by the private vendor Infinity Capital Consultants. While field-tested and funded through the Compensatory Afforestation Fund Management and Planning Authority (CAMPA) in Chhattisgarh to mitigate human-wildlife conflict, the underlying human-detection algorithms present an active design risk. When deployed generally in forest tracts, these automated systems flag any human presence as an 'unauthorized intrusion.'

Since the conventional methods to detect humans cannot differentiate between an illegal timber cutter and a legitimate Adivasi woman collecting mahua flowers, bamboo, or wood, the system inherently assumes that the FRA-protected economic activity is a threat to security. This creates a severe insufficient coordination between the two ministries. The militarized aspect of conservation zones like Bastar in Chhattisgarh due to the use of surveillance drones and thermal geofencing technology makes it hard to separate the practices linked to national security from those practiced by the indigenous forest dwellers as a part of their daily livelihood. The mechanized process of policing the process of accessing the forest area makes the rights holders trespassers.

Global Frameworks and Data Sovereignty: Why Geneva Is Watching

This conflict among domestic parties is not just an administrative malfunction or incident but a part of a wider debate that is happening across the world regarding the concept of Indigenous Data Sovereignty. This was highlighted at the 19th Session of the United Nations Expert Mechanism on the Rights of Indigenous Peoples (EMRIP), which was held in Geneva in July 2026, where international experts stressed the need to consider the impact of the rapidly growing technology of AI on the rights of Indigenous communities. This debate within EMRIP followed some important concepts that were introduced during the 60th Session of the UN Human Rights Council.

At the global level, the landmark 2024 UN General Assembly Resolution on Artificial Intelligence Governance (A/78/L.49) established a global consensus on steering AI systems toward sustainable development while strictly protecting human rights online and offline. Through encouraging coordination of their technologies with international human rights norms, the UNGA resolution reiterates that no matter how efficient a technology is, fundamental liberties should not be compromised. With regards to India, it would imply that the current mismatch between conservation surveillance technologies and tribal land rights technologies needs to be sorted out.

The Legal and Consent Gap: Collective Data under PESA and the DPDP Act

An international legal document that provides protection to indigenous peoples is known as the Convention Concerning Indigenous and Tribal Peoples, 1989 (ILO 169). About this convention, FPIC becomes the key mode of action that the state and companies need to secure before executing any project in relation to the lands of indigenous peoples. Because India has not ratified Convention 169, domestic protections remain partial and highly fragmented rather than equivalent to international FPIC standards.

This fragmentation manifests directly in India's primary domestic privacy law, the Digital Personal Data Protection (DPDP) Act, 2023. The DPDP Act is designed strictly around an individual-consent model, treating data privacy as an individual transaction between a single citizen and a data fiduciary. This individualistic framework is fundamentally incompatible with the reality of tribal data. Machine learning models trained on forest spatial data, migration patterns, and local ecological resource maps utilize collective, community-level data that describes an entire community rather than a single individual. The DPDP Act has no legal mechanism to handle, protect, or govern such collective data assets.

Furthermore, while the Forest Rights Act, 2006, vests significant democratic governance authority in the local Gram Sabha (the village assembly), this statutory body is entirely excluded from digital decisions. Because the FRA was drafted in 2006, long before the advent of automated forest surveillance, drone geofencing, and AI-enabled claimant verification, it contains no provisions addressing data governance. This has created a critical 'consent gap,' where the collective guardians of forest lands have zero legal say in how digital technologies are deployed across their territories.

A Reformist Framework: Expanding Gram Sabha Powers for Digital Governance

To resolve this institutional friction, the Indian state must introduce targeted legal reforms that formalize collective consent. The first reform requires integrating a strict, mandatory community-consent protocol directly into the design of public-sector systems like the FRA WebGIS Decision Support System. Before any spatial, demographic, or ecological data is collected, stored, or utilized for training AI models, the state must secure formal, documented consent from the local Gram Sabha. This process should mimic the consultative procedures detailed in the United Nations Development Programme's (UNDP) manual on forest and scheduled area governance, ensuring that data practices respect the local democratic structure.

The second, more structural reform is to expand the statutory powers of the Gram Sabha under the Panchayats (Extension to Scheduled Areas) Act, 1996 (PESA). Gram Sabhas in Scheduled Areas are statutory bodies established to implement Fifth Schedule principles of local self-governance. Currently, their statutory powers are confined to physical land, water, and minor forest produce. This statutory mandate must be modernized to encompass digital and data governance. Under an expanded PESA framework, the Gram Sabha should possess the explicit statutory authority to participate in all decisions regarding the deployment of digital technologies—including surveillance drones, thermal sensors, and automated conservation systems—within their territorial boundaries.

By expanding an existing, constitutionally anchored statutory institution like the Gram Sabha, the state avoids the administrative bloat of creating new regulatory bodies. Instead, it leverages PESA's robust legal framework to ensure that technology is deployed as an instrument of collaborative forest management, rather than an uncoordinated, top-down mechanism of state policing.

Violence against human rights, monitored by human-rights bodies such as the ICCA Consortium in Bastar, the expropriation of lands, and the militarization of the region are not a part of an independent narrative from the case of algorithmic disparate impact. These are precisely the conditions into which this technology is introduced without any regulation or accountability. The discussions currently taking place about Indigenous data sovereignty in Geneva's Palais des Nations have nothing abstract about them – the issue described perfectly matches what is happening in the Indian forests. It may be hard to solve the problem of militarized conservation through one law reform only, but it will solve the exact issue highlighted in this paper – a Gram Sabha having no control over the technology used in their territory. The expansion of PESA to the area of digital governance will provide the necessary legal basis for such consent.

References

1. Deokar, M. A. (2025). The Role of Gram Sabha in the Implementation of Forest Rights Act (2006) and the Achievement of Viksit Bharat 2047 Goals. Zenodo. [Link]

2. ICCA Consortium. (January 10, 2025). Alert: Adivasi communities in Bastar, India, face escalating state violence while resisting land grabs and destructive mining. [Link]

3. Infinity Capital Consultants. (2024). GAJ-DASTAK | AI-Powered Elephant Detection & Deterrence System. 

4. International Labour Organization. (1989). Convention C169 - Indigenous and Tribal Peoples Convention, 1989 (No. 169). [Link]

5. Ministry of Law and Justice. (2023). The Digital Personal Data Protection Act, 2023. Government of India. [Link]

6. Ministry of Tribal Affairs. (2006). Scheduled Tribes and Other Traditional Forest Dwellers (Recognition of Forest Rights) Act, 2006. Government of India. [Link]

7. United Nations Development Programme. (2013). Towards Creating a Model Forest and Scheduled Area Governance in Chhattisgarh: A Manual on Forest Rights Act and PESA. [Link]

8. United Nations General Assembly. (March 21, 2024). General Assembly Adopts Landmark Resolution on Steering Artificial Intelligence towards Global Good, Faster Realization of Sustainable Development. UN Press Release GA/12588. [Link]

Author is a fifth-year B.A. LL.B. (Hons.) student at Kalinga School of Law. Views are personal.