On February 19, Prime Minister Narendra Modi stood at Bharat Mandapam before French President Emmanuel Macron, UN Secretary-General Antonio Guterres, many world leaders and hundreds of leading technology executives and engineers. The occasion was the India AI Impact Summit — the first global convening of its kind held in the Global South. He spoke like a statesman: “We believe AI can benefit only when it can be shared. We should make a resolve to develop AI as a global common good. We must democratise AI. It must become a tool for inclusion and empowerment, particularly for the Global South.”
In that single statement, Modi laid down a challenge to the prevailing architecture of the AI economy — dominated by a handful of Big Tech companies, driven by proprietary models, closed datasets, and winner-take-all compute markets. What he proposed was a fundamental reframing: AI not as a strategic asset to be hoarded, but as a commons to be collectively stewarded.
To understand why this matters — and why it might actually work — one must turn not just to geopolitics, but to the Nobel Prize-winning economist Elinor Ostrom. I had argued in a paper in November 2023 at the Ostrom Workshop at Indiana University, Bloomington, US why cyberspace was a global commons. Today, AI offers a more vivid argument in that direction.
For decades, conventional wisdom held that shared resources inevitably collapse. Garrett Hardin’s “tragedy of the commons” — the idea that self-interest will deplete any shared resource — was treated as economic gospel. Ostrom demolished this orthodoxy. Through field research on Swiss mountain pastures and Maine lobster fisheries, she demonstrated that communities can manage shared resources sustainably, without privatisation or state control, provided certain institutional conditions are met: Clearly defined boundaries, rules adapted to local conditions, collective decision-making, monitoring, graduated sanctions, and dispute resolution mechanisms.
The parallels to AI are urgent. AI’s core inputs — data, algorithms, compute, and foundational models — are being rapidly enclosed by a small number of private actors. This is not merely a market outcome; it is a governance failure. When the infrastructure of intelligence itself is proprietary, its benefits accrue disproportionately to those who already hold power. The Global South, with its vast data-generating populations but limited compute and model-building capacity, is reduced to raw material for someone else’s intellectual property.
Ostrom’s framework offers a correction. If nations agree on clear governance rules — who gets access, under what conditions, how benefits are shared, how misuse is sanctioned — then AI’s foundational layers can be governed as a commons. Open-source models, shared compute infrastructure, multilingual datasets, and interoperable safety standards are all candidates for this collective stewardship. Ostrom’s key insight was that governance design matters more than the nature of the resource itself.
Modi’s MANAV vision for AI is a practical attempt to build exactly this institutional architecture. MANAV — standing for Moral and Ethical Systems, Accountable Governance, National Sovereignty, Accessible and Inclusive technology, and Valid and Legitimate Systems — is a governance philosophy anchoring AI in human values, oversight, and sovereignty.
Each pillar maps onto what Ostrom would recognise as essential conditions for commons governance. Moral and Ethical Systems create the shared normative rules without which any commons degrades. Accountable Governance provides the monitoring and sanctioning mechanisms Ostrom identified as non-negotiable. National Sovereignty — “whose data, his right” — defines the boundaries of the commons, preventing enclosure by external actors. Accessible and Inclusive technology ensures rules are made by and for all participants, not just the powerful. And Valid and Legitimate Systems build the dispute resolution and trust infrastructure that keeps the commons from fracturing.
Modi was explicit that this is not about India’s advantage alone: “Our model is to share AI development with the world and turn AI into a global common good.” The MANAV framework is being offered as a template, not a monopoly — itself a significant departure from how most powerful nations approach technology governance.
India’s credibility in making this argument rests on something real. It represents one-sixth of humanity. Its demonstrated capacity for technology-driven public goods — from UPI to CoWIN — offers a genuine proof of concept for inclusive, scalable digital infrastructure. Modi’s argument that “if an AI model succeeds in India, it can be deployed across the world” is not hubris; it is a demonstration effect backed by 1.4 billion users.
The geopolitical logic is equally compelling. The US and China are locked in an AI arms race that benefits neither the Global South nor the cause of open governance. Europe is powerful but constrained by regulatory caution. India, by contrast, offers a third way. And crucially, the coalition Modi needs is not a fantasy. France’s presence at the Summit is significant — Paris has been among the more vocal champions of “open and trustworthy AI” within the European Union. Brazil, with Lula championing digital sovereignty as part of Brazil’s G20 legacy, represents another pillar. An India-Brazil-France alignment on AI governance could be decisive within multilateral institutions, including the UN and the ITU.
With the African Union, ASEAN, and CARICOM naturally inclined toward Global South solidarity, and with a handful of progressive Global North nations willing to break from the proprietary consensus, India has both the moral authority and the coalition arithmetic to actually move the needle.
The hard work lies ahead. Declaring AI a global commons is one thing; building the institutional infrastructure to make it function as one is another. Ostrom was always clear that commons governance is not spontaneous — it requires painstaking negotiation, trust-building, and iteration. The AI commons will need agreed rules on data sovereignty, shared compute access for lower-income nations, open foundational model standards, and a legitimate multilateral body with real enforcement capacity.
India should now push for a formal Global AI Commons Treaty within the UN system, underpinned by MANAV principles, using its convening power to bring the Global South into the negotiating tent while holding Global North partners to their commitments. India, with the MANAV vision, a credible cross-hemispheric coalition, and the moral authority of the world’s largest democracy, is uniquely positioned to make the alternative real.
The writer is a defence and tech policy adviser and author of the forthcoming book The Digital Decades on 30 years of the internet in India
