- India faces a critical choice in AI governance, seeking to avoid both China's centralized, state-led control and the US's fragmented, market-driven model.
- The Global South currently lacks the capacity to shape global AI standards, risking becoming passive "norm-takers."
- India is uniquely positioned to advocate for a "third way" of regulated openness, leveraging its Digital Public Infrastructure (DPI) to prioritize public goods and inclusive growth.
Global Regulatory Approaches
- China: Employs a state-led, preventive approach characterized by algorithmic registration, mandatory pre-deployment scrutiny, and clear liability for state-aligned objectives.
- US: Relies on a market-driven, "regulate-later" approach with fragmented state-level policies, prioritizing innovation but risking platform monopolization and weak accountability.
- European Union: Utilizes a risk-based framework (EU AI Act) that imposes strict obligations on high-risk applications while maintaining democratic oversight.
Limitations of India’s Current Framework
- Existing legislation, such as the IT Act (2000) and DPDP Act (2023), focuses narrowly on data privacy and intermediary liability.
- These laws fail to address foundational generative AI risks, including model safety, algorithmic bias, lack of explainability, and the economic impact of autonomous systems.
Proposed Strategic Direction for India
- Risk-Tiered Regulation: Adopt an EU-style framework where strict ex-ante rules apply to high-risk domains like elections, biometric surveillance, and welfare allocation, while allowing experimentation in low-risk areas.
- Focus on Transparency: Prioritize public explainability and accountability rather than restrictive content control to foster innovation.
- DPI Leverage: Utilize India’s "India Stack" (e.g., Aadhaar, UPI) as a scaled testbed to demonstrate how AI can serve developmental needs, positioning India as a leader for the Global South.
- Active Engagement: Participate in international forums like WAICO to influence global standards, avoiding both isolation and uncritical alignment with existing power blocs.
This summary was generated by AI from the original article and may omit nuance or later updates. How everytldr works