On June 12, Anthropic suspended access to its frontier models, Claude Fable 5 and Claude Mythos 5, to comply with export controls imposed by the US Department of Commerce. In less than three weeks, the controls were lifted, but the episode left a message. Frontier Artificial Intelligence (AI) will now be treated much like nuclear technology: The power to turn access on or off lies with the countries building these models.

The US and China have frontier models of their own. India does not.

As a global hub for IT services and with one of the world’s largest pools of software talent, India had an early advantage. Yet it did not prepare adequately for the AI era. India’s low R&D spending, around 0.84 per cent of GDP in FY24, has not provided the foundation from which frontier AI could grow. The comparable figures were 2.6 per cent in China and 3.4 per cent in the US.…