India’s AI bet hinges on infrastructure, not just talent
The Machine Age Fund, a $1.1 billion initiative led by Andreessen Horowitz (a16z), is placing a major bet on India’s potential to develop frontier artificial intelligence companies. Despite current limitations in computing capacity and the absence of homegrown model development, Raghu Raghuram, the firm’s managing partner, argues that India’s talent pool and improving infrastructure could eventually produce its own major AI firms. According to Raghuram, startups must first build upon existing open-weight models before creating their own proprietary solutions to achieve this goal.
The fund, co-led by Raghuram and former VMware CEO Martin Casado, targets hardware and infrastructure startups. Raghuram highlighted a growing imbalance between demand and supply in AI computing needs. Companies report compute requirements growing at triple-digit rates, while supply expands at roughly double-digit levels. He emphasized that AI workloads differ fundamentally from traditional computing, creating a structural gap unlikely to close within standard industry cycles.
“The nature of the problem has changed so dramatically,” Raghuram said. “This is not something that’s going to be a five-year cycle. We are literally talking about, who knows, 30-40-year cycles.” The fund’s strategy reflects a16z’s belief that India’s AI ambitions depend on expanding infrastructure—not just software development. Raghuram acknowledged that India currently lacks frontier AI model companies but argued that the foundational knowledge to build them already exists. By now, the recipes for building good models are known, he said, adding that some day there can be big model companies from the country.
India’s AI ambitions hinge on infrastructure
Raghuram’s outlook contrasts with India’s current AI environment, where most activity revolves around services built on foreign models. His argument rests on two key assumptions. First, that India’s workforce can adapt more quickly than competitors to evolving compute demands. Second, that the rising costs and complexity of training large models will eventually open opportunities for regional players, even if they begin as infrastructure providers rather than innovators.
This perspective aligns with broader changes in India’s tech sector. Companies like Infosys are restructuring their workforce to prioritize specialized roles. The Bengaluru-based firm plans to move one-third of its 320,000 employees into niche functions, including 6,000 engineers focused on frontier technologies. CEO Salil Parekh described the shift as a response to declining employee-deployment billing rates, though Infosys has avoided the mass layoffs seen in other companies.
During its recent Americas conference, Infosys emphasized AI’s growing impact on revenue. The company reported that AI contributed 8% of total earnings in the first quarter, amounting to $1.6 billion, with double-digit quarterly growth. Infosys now competes directly with AI research labs for talent, sometimes hiring surplus staff from clients for redeployment.
Insurtech revenue under pressure from IRDAI caps
For PB Fintech, nearly 90% of operating revenue comes from insurance commissions. In fiscal year 2025–26, the company earned ₹6,089 crore from commissions, including ₹935 crore from renewals. Turtlemint, another insurtech firm, derived 20% of its ₹1,098 crore operating revenue from renewal commissions, totaling ₹225 crore in 2025–26. Banks and non-banking financial companies could also experience reduced distribution income if the caps apply retroactively to existing policies.
IRDAI’s proposed commission structure varies by product type. For individual health policies, first-year commissions would be capped at 15% for distributors and 20% for agents, with renewals limited to 5%. Portability commissions, paid when policyholders switch insurers, would be capped at 10%. The regulator’s consultation paper has prompted submissions from firms seeking clarification on whether the rules will affect renewals of policies sold before the changes take effect.
Separately, India’s deep technology ecosystem is expanding. The IIT Madras-backed Unicorn Frontier Fund-I raised ₹450 crore in its first funding round, deploying ₹55 crore across four startups. The fund’s emphasis on early-stage deeptech reflects growing investor interest in sectors like AI hardware, though scaling remains difficult due to the capital-intensive nature of these ventures.
