India missed the semiconductor train. Can it now lead in AI?

The semiconductor revolution laid the silicon foundation for today’s digital economy. Now, AI has become the next force poised to reshape industries.

For India, which missed the train on semiconductors, the AI revolution offers a second chance. Yet, challenges like limited computing infrastructure, brain drain, and a lack in deep-tech R&D – all barriers in the previous semiconductor race – may rear their heads again.

Image credit: Timmy Loen

Will India seize the AI opportunity or will history repeat itself? Will it miss the “code” just as it once missed the “chip?”

Ghosts of semiconductors past

East Asia’s semiconductor success wasn’t an accident.

Japan took the lead with government-backed research and development and strategic industry coordination in the 1970s to 1980s. Taiwan created lasting dominance through its multinational semiconductor firm TSMC’s foundry model, which now controls 70% of global foundry revenue. And while China entered late, it had a large war chest of up to US$332 billion.

Meanwhile, India – which had a promising start – faltered repeatedly.

When Fairchild Semiconductor, a founding pillar of Silicon Valley, considered setting up a packaging unit in India in the late 1950s, bureaucratic hurdles sent the firm to Malaysia instead.

India’s ambitious Semiconductor Complex Ltd (SCL), a government-owned firm that manufactures chips  and was launched in 1984, initially kept pace with global standards but was derailed by a facility fire in 1989.

See also: Who’s investing in India’s AI startups?

The critical failure wasn’t the incident itself but the eight-year delay that resulted from it – all while competitors raced ahead. This pattern of inconsistent vision, underinvestment, and bureaucratic inertia has repeatedly blocked the country’s semiconductor progress.

New frontiers, familiar challenges

The rise of AI offers India a fresh opportunity. Unlike semiconductors, AI appears more accessible, with open-source models allowing engineers in Bangalore to tinker with the same tools as their peers in Palo Alto.

The nation also enters this era with strengths: a large pool of STEM graduates, 20% of the world’s semiconductor design talent, and a thriving digital economy.

Yet, the deeper infrastructure challenges remain. Now, the bottleneck lies not in building fabs but in securing high-end computing power, which is still capital-intensive and largely foreign-controlled.

Learning from the dragon

Will history repeat itself?