‘Aggressive AI policy is the new normal’
| Photo Credit: Getty Images
The United States government recently directed American Artificial Intelligence (AI) behemoth Anthropic to suspend access to its most advanced Fable 5 and Mythos 5 models for foreign nationals on national security grounds. A U.S. Presidential order also creates a voluntary mechanism for the U.S. federal government to access such models up to 30 days before trusted partners can. The Trump administration is simultaneously considering equity stakes in leading AI companies, ostensibly to redistribute the supernormal profits expected from technological advances.These policy shifts are the latest and most dramatic in a series of sovereign actions demonstrating that governments are increasingly shaping AI policy around national advantage. Europe is moving away from its “regulate first, ask questions later” approach by investing in AI compute and promoting “Buy European” public procurement. Meanwhile, in keeping with President Javier Milei’s swashbuckling style, even Argentina aims to attract AI investment by offering a regulatory safe harbour.Navigating AI geopoliticsAggressive AI policy is the new normal. As a large IT services economy that does not have its own frontier AI systems — those requiring upwards of ten septillion floating-point operations to train — India must review global developments and articulate a coherent policy response that allows it to benefit from frontier technology without letting its economic capabilities depend on policy decisions made elsewhere.India’s IT and app companies are well positioned to scale the everyday use of AI, raising domestic productivity and competitiveness. But rapid diffusion and model dependence pull in opposite directions: using foreign AI models today is the only way to build the economic surpluses needed to depend on them less in the future. Businesses must use the best AI available to outcompete rivals. They cannot, however, manage the geopolitical risks that accompany dependence on those technologies. This is where public policy must step in. India’s discourse is stuck in a false binary between globalisation and industrial policy. India’s industries must benefit from both at the same time.The lesson is already visible in another strategic industry: Indian pharma is deeply dependent on ingredients from China, and at the same time, navigates wavering policy regimes for market access in the U.S. A Production-Linked Incentive promotes domestic manufacturing of bulk drugs but NITI Aayog’s latest assessment finds that India still sources 65% of critical ingredients from China. Industrial policies can create footholds, not instant resilience.Building strategic AI linkagesFrontier AI presents the same strategic dilemma, but on a much larger scale. India’s spends 0.6% of GDP on research and development, and the private sector accounts for a third of this. OpenAI alone projects its compute spending at $50 billion this year, over six times India’s annual private R&D spend. The implication is straightforward.India cannot outspend frontier AI investment, so it must instead deepen its backward linkages to frontier AI through government action while strengthening its forward linkages to global markets for its products and services. This starts with a whole-of-government approach, in which ministries such as external affairs coordinate closely with commerce and it, and perhaps even defence, energy, and telecom, to better serve the technology industry.Coordination alone, however, is insufficient. Firms can manage commercial risk through contracts and diversified supply chains. They cannot insure themselves against geopolitical risk or concentrated technological dependence; that is the sovereign’s role. Like export credit and hybrid-annuity infrastructure models, the government should help underwrite risks that private firms cannot efficiently bear alone. Export credit recognises that firms cannot shoulder geopolitical risks on their own and insures them against disruption. Hybrid-annuity contracts do the same for long-gestation infrastructure by having the state fund part of a project and make fixed payments over time, rather than leaving private capital to bear the full risk.Industry must step upAnd while the government can create the conditions for success, competitiveness must ultimately come from firms themselves. For its part, the Indian tech ecosystem needs to pay attention to quality and innovation that makes services and products valuable to the world.There is less room for complacency than India’s technology industry often assumes. The Philippines generates $40 billion in IT exports —already nearly a sixth of India’s IT exports — and is growing faster than the global industry.Similarly, Indian app companies have barely made a dent abroad. No Indian app features among the top 10 by downloads, in-app purchase revenue, or monthly active users. This calls for greater ambition from the industry itself.Indian technology also needs a more coherent strategic voice. Incumbent IT firms remain preoccupied with visas and market access, while start-ups are consumed by regulatory frictions and fundraising. Yet, both ultimately share the same interest: ensuring that India remains deeply connected to the world’s leading AI ecosystems while steadily building greater domestic capability.The real contest in AI is not simply over who builds the best models. It is about who captures the economic and strategic advantages they create. India’s objective should be clear: remain deeply integrated with global AI ecosystems while steadily reducing the strategic vulnerabilities that such integration creates.Vivan Sharan is Partner, Koan Advisory. Vedika Pandey is Manager, Koan Advisory Published - July 01, 2026 12:08 am IST
