Author – Piyush Chaudhary

• Abstract : 

As artificial intelligence (AI) quietly merges itself into the fabric of governance in India, a critical question comes to the forefront: Is technology making public policy more inclusive, or is it widening the gaps we’ve long struggled to close? This article investigates how AI is influencing everything from predicting droughts to distributing welfare benefits. In short, it fundamentally alters how decisions are made. India, a fast-digitizing democracy, stands at a crossroads, as on one side lies the promise of AI smarter governance, efficient public services, and data-driven solutions. On the other hand, there’s a growing concern about opaque algorithms, data privacy breaches, and AI systems reinforcing societal biases. The major aspect is, who really holds the power between the coder, policymaker, or machine. Further, this article dives into India’s evolving AI landscape from NITI Aayog’s National Strategy for AI to the upcoming AI Regulation Bill, while drawing lessons from the EU’s strict AI laws, China’s state-driven model, and the US’s corporate-learning approach. It also shines a light on the human cost,  such as how AI can either bridge or deepen inequalities, whether it’s rural farmers struggling with AI tools or marginalized communities misidentified by facial recognition software. This article doesn’t just dwell on the risks but it pushes for a future where AI serves people first through independent AI ethics bodies, transparent algorithms, and inclusive design. Thus, the objective seems very clear as AI should be a tool for public good, not a force that fuels exclusion or unchecked surveillance. In the end, it reveals that this is not just about technology,  but its  about democracy, justice, and the kind of future we want to build with AI.

• Introduction

In recent years, the integration of artificial intelligence (AI) into public policy has redefined how governments make decisions, allocate resources, and interact with their citizens. India, as a rapidly digitizing democracy, stands at the forefront of this transformation. The concept of algorithmic governance where the use of AI systems to support, guide, or even automate policy decisions is no longer a distant reality but an emerging facet of governance. From predicting drought patterns to optimizing welfare schemes, AI’s influence is becoming increasingly visible in the policymaking process (Jammu and Kashmir Policy Institute [JKPI], 2024). However, this shift raises pressing questions such as How transparent are these algorithms? And  Who holds the power between the coder designing the algorithm, the policymaker using it, or the AI itself? As India races to tackle AI’s potential, it also contends with concerns about algorithmic bias, data privacy, and the risk of excluding marginalized communities from the digital narrative. These concerns forms  the foundation of a pivotal debate such as , Can AI truly democratize public policy, or will it deepen existing inequalities? This article uncover these layers by examining India’s AI policies, its socio-economic impact, and the broader democratic implications of algorithmic governance (Pravendra Dixit, 2024).

• Understanding Algorithmic Governance and AI

Algorithmic governance is a modern approach to public administration. Where AI algorithms and machine learning models process large datasets to formulate policy decisions. Essentially, it means to use technology  not just to gather data but also to analyze and  predict outcomes, and sometimes even make decisions without direct human involvement (Dixit, 2024). This could range from AI in helping a government to  forecast crop failures and distributing subsidies accordingly. Artificial intelligence, at its core, mimics human cognitive functions like learning, reasoning, and problem-solving  through algorithms. AI operates through machine learning, neural networks, and natural language processing (NLP), allowing it to process data at an unprecedented scale (Ernst & Young [EY], 2024). In public policy, AI becomes a silent yet powerful partner, by helping in drafting welfare strategies, managing disaster responses, and streamlining public service delivery.

In India, algorithmic governance can be seen at three levels. Firstly, Decision-Support Systems (DSS) use AI to assist policymakers by offering data-driven insights. For example, during the COVID-19 pandemic, AI tools helped predict infection hotspots and guided lockdown measures in India(JKPI, 2024). Secondly, Automated Decision-Making (ADM) goes a step further by enabling AI to independently execute specific policies such as India’s direct benefit transfer (DBT) scheme, where biometric data verifies beneficiaries and facilitates real-time fund transfers (NITI Aayog, 2023). And lastly, Predictive Governance uses AI’s forecasting abilities to shape proactive policies, like anticipating unemployment spikes or food shortages (Carnegie Endowment for International Peace, 2024). A  real example is the Aarogya Setu app, an AI-powered platform that tracked COVID-19 cases and identified transmission risks, allowing the government to allocate healthcare resources strategically. Similarly, the CoWIN portal streamlined vaccine distribution by using AI algorithms in  managing registrations, ensuring fair allocation, and preventing fraud (Technology Evaluation Centre [TEC], 2024). 

• Legal and Policy Framework in India

India’s legal framework for AI is still evolving, where the government is taking cautious yet significant steps to balance innovation and regulation. The National Strategy for AI (NSAI), released by NITI Aayog in 2018, was India’s first comprehensive plan to harness AI across sectors like healthcare, agriculture, education, and smart cities. It emphasized AI’s potential to bridge socio-economic gaps while acknowledging the risks of algorithmic opacity and bias (NITI Aayog, 2023). Building on this, India is currently in the process of drafting its AI Regulation Bill. This upcoming legislation is inspired by the European Union’s AI Act. It aims to classify AI systems based on risk levels from minimal to high-risk, ensuring that critical AI applications like biometric surveillance undergo strict scrutiny. The bill seeks to create accountability mechanisms, mandating transparency in AI algorithms used for public services (Carnegie Endowment, 2024).

Additionally, the Data Protection Act, 2023 provides a primary layer for AI governance by regulating data collection, consent, and usage. Since AI systems rely heavily on large datasets, This law has safeguarded citizens’ data privacy, ensuring that AI applications, whether in law enforcement or public welfare do not infringe on individual rights (Access Partnership, 2024). The Technology Evaluation Centre (TEC) also released an AI Policy Paper (2024), which advocates for algorithmic transparency. It also proposes mandatory AI audits for government projects and stresses the need for explainable AI (XAI) , where AI models provide clear, human-understandable explanations for their decisions (TEC, 2024). However, despite these measures, gaps are still existing. Experts point out that the absence of a dedicated AI regulatory authority in India is making it difficult to oversee AI-related breaches or biases comprehensively (Dixit, 2024). There is also growing concern over ‘regulatory capture’ where big tech firms could influence AI policies to serve corporate interests over public welfare.

• Political and Democratic Implications of AI in Governance

The introduction of AI into governance is not just a technological shift, but  it’s a deeply political one as AI systems have increasingly influenced public policy in India. They raised urgent questions about democratic accountability, transparency, and the balance of power between the state, corporations, and citizens. One of the most critical concerns is algorithmic opacity, the “black box” nature of AI models, where even developers struggle to explain how certain decisions are made. In a democracy like India, where transparency and public participation are constitutional values, opaque AI systems pose a significant threat. For instance, AI-driven facial recognition technology (FRT) is now used by law enforcement in cities like Delhi and Hyderabad to track criminal activity. However, reports show that these systems have disproportionately misidentified individuals from marginalized communities, raising concerns about systemic biases embedded in AI algorithms (Dixit, 2024). Without clear oversight, such practices risk eroding civil liberties under the guise of technological advancement.

Another issue is the centralization of power. AI algorithms often rely on vast amounts of data, and control over this data increasingly rests with the state and large tech companies as the collaboration between government platforms like Aarogya Setu and private tech firms during the pandemic highlighted the delicate balance between public health management and data privacy. Critics argue that AI-enhanced surveillance, if it’s unchecked, could lead to “digital authoritarianism,” where technology becomes a tool for mass surveillance rather than public welfare (Carnegie Endowment, 2024).

Furthermore, AI’s influence on electoral processes is growing. Political parties now deploy AI-powered data analytics to micro-target voters, tailoring campaigns based on their social media activity, location, and browsing history. While this might enhance campaign efficiency, it raises ethical concerns about voter manipulation and the spread of misinformation. The lack of vigorous AI regulations in electoral processes means that AI-driven propaganda could easily be weaponized to polarize public opinion (JKPI, 2024). To uphold democratic integrity, it is vital for AI systems used in governance to be transparent, explainable, and accountable. Establishing independent AI ethics committees and mandating algorithmic impact assessments can help mitigate the risks posed by AI in India’s political landscape.

• Socio-Economic Consequences of AI-Driven Public Policy

The socio-economic impact of AI in governance is a double-edged sword. On the one hand, AI presents immense opportunities to enhance welfare delivery, improve public infrastructure, and boost economic growth. On the other hand, it risks deepening inequalities if not implemented with a people-centric approach. In sectors like healthcare, AI has already demonstrated its transformative power. For instance, AI-powered diagnostic tools have improved early detection of diseases in rural areas, where access to specialist doctors is limited. Projects like the NITI Aayog sponsored “AI for All” initiative aim to leverage AI for predictive healthcare, reducing the burden on an already stretched public health system (NITI Aayog, 2023). However, these benefits remain unevenly distributed. Studies show that AI-based health interventions often bypass remote tribal areas due to inadequate digital infrastructure, highlighting the “digital divide” in AI deployment (AIM Research, 2024).

In agriculture, AI has been used to predict crop yields and optimize resource distribution. Startups like CropIn use AI models to advise farmers on pest control and water management, by potentially boosting productivity. Yet, small farmers with limited digital literacy struggle to access these AI tools, which further widened the gap between tech-savvy agribusinesses and traditional farming communities (TEC, 2024). Employment is another contentious issue. AI automation has the potential to streamline bureaucratic processes for example, using AI chatbots to handle routine queries on government portals like UMANG but it also threatens job displacement. A report by the Carnegie Endowment (2024) warns that AI adoption in sectors like manufacturing and public services could disproportionately affect low-skilled workers, exacerbating unemployment. Without reskilling programs, AI could entrench economic disparities rather than bridging them.

Furthermore, algorithmic biases often replicate existing societal prejudices. An AI system used in an Indian bank’s loan approval process was found to reject applicants from lower-income areas more frequently, as the algorithm had been trained on historical data that reflected systemic bias. This highlights if AI is left unchecked, how it can reinforce discrimination rather than eliminate it (Dixit, 2024). To ensure AI serves social equity, India must integrate inclusive design principles into AI policies, ensuring marginalized groups have a say in how these technologies are deployed.

• Comparative Analysis of  India with Global AI Governance Models

India’s AI governance journey can be better understood through a comparative lens, by examining how other nations balance innovation and regulation. The European Union’s AI Act stands out as a benchmark for AI regulation. The EU classifies AI systems into risk categories  minimal, limited, high, and unacceptable  imposing strict obligations on high-risk AI applications like facial recognition and credit scoring. India’s proposed AI Regulation Bill draws inspiration from this risk-based approach, aiming to hold AI systems accountable while fostering innovation (Carnegie Endowment, 2024).

In contrast, China’s AI governance focuses on state control and AI-driven surveillance. The Chinese government uses AI for predictive policing and social credit systems, raising global concerns about privacy violations. While India’s democratic setup prevents such extreme measures, AI surveillance in cities like Delhi shows how unregulated AI tools can quietly encroach on civil liberties if left unchecked (JKPI, 2024).

The United States follows a market-driven AI approach, with tech giants like Google and Microsoft shaping AI policies through public-private partnerships. However, the lack of comprehensive federal AI laws in the US highlights the risks of corporate influence, a concern India must address by ensuring AI regulations are driven by public interest rather than corporate lobbying (Dixit, 2024). India’s AI strategy, therefore, stands at a crossroads choosing between fostering innovation and safeguarding democracy. Striking a balance will require not just importing global best practices but tailoring AI governance to India’s unique socio-political context.

• Policy Recommendations

The first and most pivotal step is to establish an Independent AI Ethics Authority (AIEA). This body should operate autonomously, free from corporate and political influence, to oversee AI deployment in public policy. The AIEA would be responsible for conducting AI audits, investigating algorithmic bias complaints, and formulating ethical AI guidelines. According to the Carnegie Endowment for International Peace (2024), independent oversight is critical to prevent AI systems from reinforcing systemic inequalities or being misused for mass surveillance. To bring this to life, the government must pass an AI Governance Act, legally constituting the AIEA under parliamentary oversight. The authority’s composition should reflect diversity  AI experts, ethicists, legal professionals, and civil society representatives  to ensure inclusive governance. Additionally, the AIEA must publish annual reports on AI usage, ensuring public accountability. A timeline of 12–18 months for drafting legislation and 6 months for operational setup would provide a solid foundation for this regulatory body (Dixit, 2024).

Secondly, Algorithmic Impact Assessments (AIA) should be made mandatory for all AI systems used in public administration. AIAs would evaluate the social, economic, and ethical consequences of AI projects before their implementation. This aligns with the European Union’s AI Act, which categorizes AI systems based on risk levels and mandates preemptive assessments for high-risk AI tools (Carnegie Endowment, 2024). To operationalize this in India, the Data Protection Act, 2023 should be amended to incorporate AIAs as a legal requirement. The impact assessments must cover critical areas such as bias detection, privacy safeguards, and potential harm to marginalized communities. As a starting point, the government can conduct pilot AIAs on existing AI-powered initiatives like Aarogya Setu and CoWIN to identify flaws and build an adaptable framework (Technology Evaluation Centre [TEC], 2024). Transparency will be key; AIAs should be publicly accessible, allowing civil society and watchdog groups to scrutinize government AI projects (JKPI, 2024).

The third recommendation focuses on inclusive AI development to prevent algorithmic bias and foster equitable technological growth. AI systems often replicate social inequalities when they are trained on biased datasets, as seen in cases where AI models disproportionately denied bank loans to individuals from marginalized backgrounds (Dixit, 2024). To address this, the government should co-design AI tools in collaboration with grassroots organizations, farmer unions, and gender rights groups. Initiatives like AI Innovation Labs in rural districts can support the development of localized AI solutions, such as predictive models for crop diseases or maternal health tracking systems. Accessibility must be central AI interfaces should be available in regional languages and tailored for people with limited digital literacy (NITI Aayog, 2023). Additionally, establishing a Community Oversight Board composed of activists, AI researchers, and community leaders can ensure AI projects align with public interest.

Furthermore, AI literacy and reskilling programs are vital to bridging the digital divide. While AI promises efficiency, it also threatens job displacement, particularly in sectors like manufacturing and public services (Carnegie Endowment, 2024). To counter this, AI education should be integrated into national programs like Skill India and Digital India. These programs can offer sector-specific AI modules: public servants can learn about AI ethics and algorithmic accountability, farmers about AI-powered crop predictions, and youth about coding and machine learning. The creation of a National AI Fellowship could further empower grassroots innovators to design AI solutions for their communities. To extend AI education to remote areas, Mobile AI Labs  set up in collaboration with tech firms can offer hands-on training and real-time support (AIM Research, 2024).

Finally, ensuring data transparency and algorithmic explainability is non-negotiable. AI systems used in governance must disclose how their algorithms process data, make decisions, and mitigate risks. According to the Technology Evaluation Centre (2024), the proposed AI Regulation Bill should mandate Explainable AI (XAI) practices, where AI models provide human-understandable justifications for their outputs. This would be particularly pivotal for AI systems used in welfare programs, law enforcement, and credit scoring. Data transparency laws should compel both government agencies and private AI developers to release detailed documentation on how their algorithms function, minimizing the risk of “black box” AI models (Access Partnership, 2024). Public consultation processes should also be embedded in AI policy formulation, inviting feedback from citizens, technologists, and activists to foster participatory AI governance.

• Conclusion 

As artificial intelligence steadily reshapes public policy in India, one thing is clear as AI is not just a technological tool; but  it is a political force. It has the power to make governance more efficient, responsive, and data-driven, but it also carries the risk of reinforcing existing inequalities, centralizing power, and eroding democratic accountability. India stands at a pivotal juncture. While initiatives like NITI Aayog’s National Strategy for AI (NITI Aayog, 2023) and the upcoming AI Regulation Bill inspired by the European Union’s AI Act (Carnegie Endowment, 2024)  signal the government’s intent to harness AI for public good, the road ahead is far from simple. The challenge is not merely about adopting AI  it’s about ensuring that AI works for everyone. For AI to truly serve democracy, it must be transparent, explainable, and inclusive. Algorithms used in public policy cannot operate as “black boxes”  ; their decision-making processes must be open to scrutiny. The establishment of an Independent AI Ethics Authority (Carnegie Endowment, 2024), mandatory Algorithmic Impact Assessments (AIA) (Technology Evaluation Centre, 2024), and the integration of marginalized voices into AI development (Dixit, 2024) are not just policy suggestions, they are necessities. Ultimately, the future of AI governance in India will be shaped by the choices we make today. Will AI be a tool for empowerment or exclusion? Will it bridge the digital divide or deepen it? The answer lies not in the algorithms themselves, but in the people who design, regulate, and challenge them. In this rapidly evolving landscape, one truth remains: AI should serve humanity not the other way around. If India can strike a balance between innovation and accountability, it has the potential to craft a model of algorithmic governance that is not only technologically advanced but also just, inclusive, and deeply democratic.

REFERENCES

1. Access Partnership. (2024). The key policy frameworks governing AI in India. Access Partnership. https://accesspartnership.com/the-key-policy-frameworks-governing-ai-in-india/
2. AIM Research. (2024). The role of artificial intelligence in enhancing public policy. AIM Research. https://aimresearch.co/council-posts/the-role-of-artificial-intelligence-in-enhancing-public-policy
3. Carnegie Endowment for International Peace. (2024). India’s advance on AI regulation. Carnegie Endowment. https://carnegieendowment.org/research/2024/11/indias-advance-on-ai-regulation
4. Dixit, P. (2024). Navigating the development and governance of AI ecosystems in India: Challenges, opportunities, and policy recommendations. ResearchGate. https://www.researchgate.net/publication/383943937_Navigating_the_Development_and_Governance_of_AI_Ecosystems_in_India
5. Ernst & Young (EY). (2024). Is generative AI beginning to deliver on its promise in India? EY. https://www.ey.com/content/dam/ey-unified-site/ey-com/en-in/services/ai/documents/ey-is-generative-ai-beginning-to-deliver-on-its-promise-in-india-aidea-of-india-update.pdf
6.  European Union. (2024). The AI Act: Regulating artificial intelligence in the EU. European Commission.
7. Jammu and Kashmir Policy Institute (JKPI). (2024). AI’s role in shaping public policy and governance: The path forward. JKPI. https://www.jkpi.org/ais-role-in-shaping-public-policy-and-governance-the-path-forward
8.  Kashik, K. (2024). Algorithmic governance in India: Evaluating AI’s role in public policy. National Centre for Good Governance. https://ncgg.org.in/sites/default/files/lectures-document/Kshitija_Kashik__Research_Paper.pdf
9. NITI Aayog. (2023). National strategy for artificial intelligence. NITI Aayog. https://www.niti.gov.in/sites/default/files/2023-03/National-Strategy-for-Artificial-Intelligence.pdf
10. Pravendra, D. (2024). Navigating the development and governance of AI ecosystems in India: Challenges, opportunities, and policy recommendations. ResearchGate.
11. Springer. (2024). AI and public policy: Challenges and solutions. Springer. https://link.springer.com/chapter/10.1007/978-3-031-68326-8_7
12. Tandfonline. (2024). AI in public policy: Bridging governance and technology. Taylor & Francis. https://www.tandfonline.com/doi/full/10.1080/22041451.2024.2346428
13. Technology Evaluation Centre (TEC). (2024). AI policies in India: A status paper. TEC. https://www.tec.gov.in/pdf/Studypaper/AI%20Policies%20in%20India%20A%20status%20Paper%20final.pdf
14. Wiley Online Library. (2024). AI and its role in algorithmic governance. Wiley. https://onlinelibrary.wiley.com/doi/abs/10.1111/soc4.12955
15. Wiley Online Library. (2024). AI regulations: A comparative study. Wiley. https://onlinelibrary.wiley.com/doi/abs/10.1111/rego.12367
16. ProQuest. (2024). Algorithmic governance in India: Evaluating AI’s role. ProQuest. https://search.proquest.com/openview/3e2a1b4386b7d1b63a3cc6e23ea355b9/1?pq-origsite=gscholar&cbl=2026366&diss=y
17. Google Books. (2024). Algorithmic governance in India: Evaluating AI’s role in shaping public policy. Google Books. https://books.google.com/books?id=aCcHEQAAQBAJ
18. OAPEN Library. (2024). AI and public governance: Global and local perspectives. OAPEN. https://library.oapen.org/bitstream/handle/20.500.12657/23985/1006150.pdf
19. Emerald Insight. (2024). Developing AI policies: A roadmap for India. Emerald Publishing. https://www.emerald.com/insight/content/doi/10.1108/dprg-10-2024-0272/full/htm
20. MDPI. (2024). AI, public policy, and the future of governance. MDPI. https://www.mdpi.com/2078-2489/15/9/556