The United States government recently instructed American AI company Anthropic to restrict access to its advanced Fable 5 and Mythos 5 models for foreign nationals due to national security concerns. A presidential order also establishes a voluntary process allowing the U.S. federal government to review such models up to 30 days before trusted partners. The Trump administration is also weighing equity stakes in major AI firms to share expected profits from technological progress.
These changes reflect a broader trend of governments aligning AI policy with national interests. Europe is shifting from its initial regulatory focus by investing in AI infrastructure and favoring European suppliers in public contracts. Argentina, under President Javier Milei, seeks to draw AI investment through a regulatory safe harbor.
Navigating AI geopolitics
Strong AI policies are now standard. As a major IT services economy without its own frontier AI systems, which require massive computing power to train, India needs to assess global trends and develop a clear policy. This approach should enable gains from advanced technology while avoiding reliance on decisions made abroad.
India’s IT and app firms can expand everyday AI use to boost productivity. Yet quick adoption and dependence on foreign models create conflicting pressures. Companies need the best available AI to compete, but they cannot handle associated geopolitical risks alone. Public policy must address this gap. India’s debate often presents a false choice between globalization and industrial policy; both can support industry simultaneously.
Similar patterns appear in pharmaceuticals, where India depends on Chinese ingredients while facing changing U.S. market rules. Production incentives aim to increase local manufacturing, yet India still imports most critical components from China. Such policies build initial capacity rather than immediate self-reliance.
Building strategic AI linkages
Frontier AI poses comparable challenges on a larger scale. India allocates 0.6 percent of GDP to research and development, with the private sector contributing one-third. OpenAI alone plans $50 billion in compute spending this year, exceeding six times India’s annual private R&D outlay.
India cannot match frontier AI investment levels. Instead, it should strengthen connections to advanced AI through government measures and improve access to global markets. This requires coordinated efforts across ministries including external affairs, commerce, IT, defense, energy, and telecom.
Coordination alone is not enough. Firms handle commercial risks via contracts and supply diversification but cannot manage geopolitical or technological dependence risks. Government support is needed, similar to export credit or hybrid-annuity infrastructure models that share risks private entities cannot bear alone.
Industry must step up
Government can set favorable conditions, but firms must deliver competitiveness. India’s tech sector should focus on quality and innovation valued globally.
Complacency carries risks. The Philippines already earns $40 billion from IT exports, nearly one-sixth of India’s total, with faster growth. Indian apps also have limited international reach, with none ranking in top 10 lists by downloads or revenue. Greater ambition is required from the industry.
