Picture an Indian AI company that builds a competitive model, raises funding and wins customers. Then a government decision puts a critical part of its business out of reach. Anuraag Saxena, a public policy leader, says Indian AI firms underestimate this risk. He spoke at Cypher 2026 in Bengaluru.
Sovereignty Is Not Self-Sufficiency
Speaking at the event, Saxena said AI sovereignty does not mean building everything at home. It means no outside power can unilaterally disrupt a company’s ability to operate. In his words, sovereignty means no one else can unilaterally switch you off.
For AI firms, that could mean losing advanced computing infrastructure, critical materials, skilled workers or key markets. Decisions made far outside their control could trigger any of these.
Natural Resources Return as a Limit
Technology once helped overcome geographic limits, Saxena said. That edge is now under pressure, as access to natural resources increasingly decides what countries and companies can build. He called it a basic macroeconomic trend that has received too little attention.
Rare-earth magnets show the risk. A September 2026 report by Boston Consulting Group and the Automotive Component Manufacturers Association of India found that India sources nearly 85% of its rare-earth magnets from China, Autocar Professional reported. China’s export curbs, introduced in October 2025, were due to expire on November 10, 2026.
AI infrastructure faces the same exposure. Access to advanced chips and computing capacity can depend on export controls and a small group of suppliers.
Talent and Immigration Politics
Talent is another weak point. On October 8, 2026, the Trump administration suspended labour certification applications for eight technology companies, including TCS and Infosys. The move showed how immigration policy can upend workforce plans and employee mobility.
Saxena pointed to a contradiction. Companies want skilled engineers, while voters want tighter immigration controls. Immigration is an electoral issue, he said. For firms working across borders, this creates competing regulations. Walls often go up for electoral reasons, he added. Political decisions need not target technology to hurt a company’s ability to operate.
Geopolitical Insurance
Saxena urged AI founders to look beyond funding, compute costs, hiring and customers. They should also ask whether any government could restrict a resource or market they depend on. A sudden policy change could turn off a tap and halt a business.
His advice was to think geopolitically from day one. He called redundancy a form of geopolitical insurance. Companies should hedge across jurisdictions and avoid relying on a single provider or vendor. In practice, that could mean using several compute providers, checking exposure to export controls and preparing for regulatory change in key markets.
The aim is not to remove every international dependency, which no country could do. It is to find the critical ones and build alternatives before a disruption hits.
Lessons From the 1990s
Saxena doubted that India can grow its AI industry on domestic demand alone. That would work only if other countries saw India mainly as a consumer market. The US and China treat AI as a tool of technological dominance, he said. His view: Shameless national interest should come first.
He recalled the 1990s, when Indian firms such as Zenith and HCL sold 286 computers at about half the price of Dell and HP. India still failed to build lasting hardware leadership. He said India went on to lose the software, enterprise solutions and social media contests too, and that he does not want it to lose the AI one.
For Saxena, the next decade will favour companies that understand the geopolitical conditions their AI businesses depend on, not just those with the best models. India must build domestic capability while accepting that sovereignty is not self-sufficiency. The test is whether Indian AI firms can keep building, deploying and selling when geopolitical priorities shift.
