In a bold call to action at the India AI Impact Summit 2026, Amitabh Kant, ex-CEO of NITI Aayog, declared that India and fellow Global South countries must forge their own path in AI. The culprit? Dominant tech firms exploiting data from emerging markets to supercharge their LLMs, while offering little in return.
Kant revealed India generates over a third more data than the US for these models. This data goldmine from the Global South is refining cutting-edge AI, but the profits and control remain with a handful of giants. They risk packaging these insights into expensive products, pricing out the very contributors.
Kant’s solution is straightforward: create indigenous LLMs grounded in local data. This ensures equitable AI access, making it cheap, responsible, and available in multiple languages. Such models could revolutionize daily life, bridging divides rather than deepening them.
He painted a worrying picture of AI’s double-edged sword. Massive investments are accelerating progress, but without safeguards, societies could fracture further. The real test lies in deploying AI for the underprivileged—improving schooling, health metrics, and nutrition standards.
‘What was once physically unfeasible is now achievable through AI,’ Kant noted optimistically. Yet, he issued a stern warning: sideline the poor, and inequalities will skyrocket. Education and skill-building must be AI’s priorities.
Kant’s vision challenges the status quo, pushing for AI that serves the Global South’s 4 billion people. It’s time for data ownership to translate into real empowerment, not exploitation.