Contents
Introduction
India’s AI governance is evolving through IT Rules, financial regulation, and data protection law, even as it ranks among top AI adopters but lags behind the US–China axis in frontier model development.
Existing AI Regulatory Framework in India
- IT Act and IT Rules, 2021: India regulates AI indirectly by imposing due diligence obligations on intermediaries, mandating removal of unlawful content, curbing deepfakes, and requiring labelling of synthetically generated content.
- Data Protection Regime: The Digital Personal Data Protection Act, 2023 embeds principles of lawful processing, purpose limitation, and accountability, indirectly governing AI systems that rely on personal data and automated decision-making.
- Sectoral Regulation:
- RBI: Introduced model risk management expectations and the FREE-AI framework to ensure explainability, fairness, and governance in AI-driven credit systems.
- SEBI: Mandated accountability for AI tools used by regulated entities, focusing on auditability and human oversight.
- Regulatory Character: India’s approach remains reactive and fragmented, relying on existing laws rather than a comprehensive AI-specific duty of care or product safety regime, especially for psychological and consumer harms.
Comparative Perspective and Regulatory Gaps
- Absence of AI Consumer Safety Framework: Unlike China’s draft rules on emotionally interactive AI, India lacks explicit obligations addressing psychological dependence, behavioural manipulation, or algorithmic harm.
- Trade-off with Intrusiveness: While China’s model risks excessive surveillance, India’s lighter-touch approach risks regulatory incompleteness, especially in high-risk AI applications such as recommender systems, fintech, and health-tech.
- EU AI Act Contrast: The EU follows a risk-based regulation, categorising AI systems into unacceptable, high-risk, and minimal-risk, offering India a template without stifling innovation.
Innovation Constraint: India’s Resource Deficit
- Computational Access: India lacks affordable access to high-performance computing and GPUs, a critical bottleneck in training large language and foundation models.
- R&D and Frontier Models: India is a major AI adopter but not a frontier model builder, increasing dependency on foreign, privately owned models.
- Public Investment Gaps: Compared to China’s state-backed AI compute clusters and the US CHIPS–AI ecosystem, India’s public procurement and mission-mode funding remain limited.
Workforce Upskilling: Strategic Imperative
- Human Capital Advantage: With the world’s largest STEM workforce, India can convert demographic scale into AI leadership through skilling, reskilling, and interdisciplinary AI ethics education.
- Policy Initiatives: Programs like IndiaAI Mission and Digital India can integrate AI training across governance, industry, and academia.
- Bridging Research–Industry Gap: Translating academic AI research into deployable products can reduce reliance on imported models and strengthen domestic innovation.
Balancing Innovation with Ethical Governance
- Downstream Regulation: India should regulate high-risk AI use cases, not upstream model development, by imposing obligations like incident reporting, algorithmic audits, and human-in-the-loop safeguards.
- Responsible AI Principles: Embedding fairness, transparency, explainability, and accountability aligns with Supreme Court jurisprudence on privacy and dignity under Justice K.S. Puttaswamy (2017).
- Avoiding “Regulate First, Build Later” Trap: Overregulation without domestic capacity may deepen technological dependency rather than sovereignty.
Conclusion
As technology must serve constitutional values. Echoing the Economic Survey, India’s AI future hinges on capacity-building, not control—innovation tempered by ethical governance.
