Overview of AI’s Impact on Scheduled Castes and Scheduled Tribes in India
SR Darapuri, National President, All India Peoples Front
(Asian independent) Artificial Intelligence (AI) is poised to have a profound, dual-edged impact on India’s Scheduled Castes (SCs) and Scheduled Tribes (STs), who together comprise about 25% of the population and have historically faced socioeconomic marginalization due to caste hierarchies, limited access to resources, and systemic biases. On one hand, AI offers opportunities for empowerment through enhanced education, land rights management, and economic inclusion. On the other, it risks exacerbating inequalities by perpetuating caste biases embedded in data, widening the digital divide, and enabling discriminatory practices in sectors like justice, healthcare, and employment. These impacts are influenced by India’s rapid AI adoption—projected to contribute $450-500 billion to GDP by 2025—amidst uneven digital infrastructure and historical data skewed toward dominant castes. While some stakeholders, including government initiatives like Digital India, view AI as a tool for inclusive growth, civil society and academic sources highlight risks of reinforcing colonial-era oppressions, such as those from the Criminal Tribes Act. Balanced perspectives suggest that without targeted interventions, AI could widen disparities, but with ethical frameworks, it might help dismantle caste barriers.
Positive Impacts
AI has the potential to uplift SC/ST communities by addressing longstanding barriers in education, land rights, and economic opportunities. For instance, in education, AI-driven tools can provide multilingual instruction and personalized learning, crucial for SC/ST students who often face language barriers and higher dropout rates. Platforms under initiatives like PM eVidya use AI for early warning systems to reduce dropouts among at-risk SC/ST youth, potentially improving access in remote tribal areas. In land rights, AI enables efficient mapping and claim verification under the Forest Rights Act, helping ST communities resolve disputes and monitor encroachments via satellite imagery and machine learning. This could empower tribes in states like Odisha and Maharashtra by digitizing records and promoting transparency.
Economically, AI supports micro-enterprises in tribal regions, fostering entrepreneurship and sustainable development in states like Odisha and Andhra Pradesh. Optimistic views suggest AI could annihilate caste by predicting violence, developing preventive measures, and creating equal-opportunity platforms, potentially reducing systemic biases and bringing “second independence” for lower castes. Grassroots AI tools like Kisan AI, developed by Adivasi (ST) communities, create counter-narratives in local languages to challenge state biases and overturn illegal land grabs.
Sector Positive Impact on SC/ST Examples
* Education *Personalized, multilingual learning reduces dropouts and bridges gaps. *AI platforms analyzing student data for tailored content; assistive tech for disabilities. |
*Land Rights * Efficient mapping and dispute resolution. *AI under Digital India Land Records Modernisation Programme monitors encroachments.
* Economy * Entrepreneurship and micro-enterprise empowerment. * AI tools for tribal businesses in rural areas.
*Social Justice * Predicting and preventing caste violence. * AI platforms for equal opportunities and bias reduction.
Negative Impacts
Conversely, AI often reinforces caste hierarchies due to biased training data drawn from colonial and dominant-caste sources, leading to discriminatory outcomes. In policing and justice, systems like Trinetra and SUPACE target SC/ST areas as “high-risk” 73% more often, using biased data to criminalize communities and dismiss cases like sexual assaults against Dalit women. The digital divide exacerbates this: SC/ST groups have lower internet access (e.g., 71% for ST households vs. 92% for upper castes), limiting AI benefits and widening skills gaps due to education and income disparities.
In healthcare and employment, AI tools misdiagnose diseases in SC/ST due to skewed data, while large language models (LLMs) discriminate based on caste-associated names or accents, excluding them from jobs, loans, and services. Humanitarian mapping shows biases affecting ST in disaster response. Linguistic exclusion (AI favoring English) and data privacy risks further marginalize tribes, eroding cultural practices.
Sector Negative Impact on SC/ST | Examples
* Justice & Policing *Biased surveillance and criminalization. * AI marking Dalit villages as high-risk; dismissing assault cases
*Digital Access *Widened divide in skills and opportunities. * Lower internet and ICT access perpetuates exclusion.
* Healthcare & Employment *Discriminatory diagnostics and exclusions. * Misdiagnosis; bias in job/loan algorithms based on names/accents.
* Culture & Privacy *Erosion of traditions and data risks. * AI disrupting tribal practices; privacy infringements.
Mitigation Strategies and Future Outlook
To mitigate negative impacts, stakeholders advocate for decolonizing AI data, conducting caste audits, and embedding constitutional protections like Article 350 for linguistic justice. Involving SC/ST scholars in AI development, promoting diversity in tech, and updating laws like the Digital India Act to include anti-caste clauses are key. Community-led initiatives, capacity building, and ethical guidelines under NITI Aayog’s principles can ensure AI empowers rather than oppresses. If implemented, AI could shift from perpetuating hierarchies to fostering equity; otherwise, it may deepen divides, as seen in current trends./
