Opinion India’s AI governance blueprint is Global South’s best bet. Here’s why

Second, six pillars of practical recommendations span infrastructure, capacity building, policy and regulation, risk mitigation, accountability, and institutions. This horizontal integration ensures that AI governance isn’t siloed within technology ministries but permeates across sectors — from finance to healthcare, education to agriculture.

Third, an action plan with clear timelines breaks the ambitious vision into achievable milestones. Short-term priorities include establishing governance institutions and developing India-specific risk frameworks. Medium-term goals focus on standardisation and legal amendments. Long-term objectives emphasise continuous review and future-ready legislation — a refreshing acknowledgement that AI governance cannot be static.

Fourth, practical guidelines for industry and regulators translate lofty principles into actionable steps. Industry players must comply with existing laws, adopt voluntary frameworks, and establish grievance mechanisms. Regulators, meanwhile, must prioritise real harms over hypothetical risks and avoid compliance-heavy regimes that stifle innovation.

India’s approach is revolutionary precisely because it’s evolutionary. Rather than crafting an entirely new regulatory apparatus from scratch — the European Union’s path with its AI Act — India leverages existing legal frameworks. The Information Technology Act, Digital Personal Data Protection Act, Consumer Protection Act, and sector-specific regulations already cover many AI-related harms. The framework simply clarifies how these apply to AI systems and identifies genuine gaps requiring targeted amendments.

This pragmatism will resonate strongly with developing nations facing similar constraints: Limited regulatory capacity, nascent AI ecosystems, and the pressing need to balance innovation with safety. India’s emphasis on voluntary measures — industry codes, technical standards, self-certifications — supported by appropriate incentives offers a viable middle path. Countries in the Global South, representing a majority of the world’s population, need governance models that enable rather than constrain AI adoption. India’s framework provides exactly that blueprint.

The framework’s techno-legal approach represents another exportable innovation. By embedding regulatory requirements directly into system architecture—what the document calls “compliance-by-design” — India demonstrates how Digital Public Infrastructure can be leveraged for governance at scale. The proposed Data Empowerment and Protection Architecture (DEPA) for AI Training exemplifies this: It enables privacy-preserving mechanisms during model development while maintaining transparency and auditability. For nations building digital public infrastructure from the ground up, this integrated approach offers tremendous value.

Perhaps most significantly, India’s framework operationalises the vision of “AI for All” articulated by Prime Minister Narendra Modi. The emphasis on multilingual interfaces, accessibility in Tier-2 and Tier-3 cities, and AI-driven solutions for agriculture, rural healthcare, and vernacular education speaks to India’s unique demographic diversity and development challenges. This inclusive lens — ensuring AI benefits reach “the last citizen” — offers a powerful counter narrative to narratives of AI as an elite technology amplifying existing inequalities.

The establishment of the AI Safety Institute (AISI) and the proposed AI Governance Group (AIGG) creates an institutional architecture that balances technical expertise with policy coordination. The AISI’s mandate includes research, risk assessment, safety testing, and international collaboration — positioning it as India’s authoritative voice on AI safety issues. Meanwhile, the AIGG’s whole-of-government approach ensures coordination across ministries and sectoral regulators, avoiding the fragmentation that has plagued AI governance efforts elsewhere.

That said, implementation will test the framework’s true mettle. The success of voluntary measures depends on creating genuine incentives for compliance — whether through access to regulatory sandboxes, public recognition, technical assistance, or venture capital preferences. Without these carrots, principles risk remaining aspirational.

The framework’s stance on copyright and AI training — awaiting recommendations from a separate DPIIT committee — highlights one unresolved tension. While the document acknowledges the need for text and data mining exceptions to enable AI development, balancing innovation with creator rights remains contentious globally. India’s solution here will be closely watched.

Similarly, the emphasis on “India-specific risk assessment frameworks” and empirical evidence of harm is prudent but time-consuming. The proposed AI incidents database will require several years to generate actionable insights. In the interim, regulators must navigate without the comprehensive evidence base the framework envisions.

The AI Action Summit represents India’s opportunity to showcase this framework on the world stage. Unlike narrow safety-focused summits or industry-dominated convenings, India can position itself as the champion of balanced, inclusive, and implementable AI governance. The framework’s emphasis on leveraging Digital Public Infrastructure, enabling innovation in resource-constrained environments, and prioritising societal benefit over corporate interests will resonate across Asia, Africa, and Latin America.

India’s timing is impeccable. As the European Union’s AI Act enters implementation — with its complexity drawing criticism — and as the United States maintains its pro-innovation, light-touch approach, the world needs alternatives that combine elements of both. India’s framework, with its risk-based approach that doesn’t classify AI systems into rigid categories, voluntary measures supported by clear enforcement of existing laws, and institutional coordination without creating new regulatory bureaucracies, offers that middle path.

The true test of these guidelines will unfold over the coming years as they are implemented across sectors, refined based on real-world experience, and potentially adopted or adapted by other nations. But as a statement of intent and a comprehensive governance blueprint, the IndiaAI Governance Guidelines represent India’s most significant contribution yet to the global conversation on responsible AI development.

Bhattacharjee is a defence and tech policy adviser and former country head of General Dynamics. Bal is an advocate, and parliamentary and legislative researcher