India’s AI ambitions in national security will falter without scientific temper, institutional reform and a retreat from sectarian politics
Artificial Intelligence (AI) has been the subject of many brainstorming sessions by different interest groups, ranging from trade unions anxious about labour-saving technologies that displace workers, and investor communities fretting over a possible AI bubble burst, to industry groups collectively chewing their nails over potential loss of competitiveness in the global marketplace.
A recent such session in Delhi sought to explore the implications of AI for national security and arrive at some consensus on what ought to be done.
AI's impact on national
An initiative jointly undertaken by think tanks NatStrat and the Strategic Foresight Group, and facilitated by media platform Founding Fuel, the session drew attention to the many ways in which AI would impact national security, with potential impact on the nation’s scientific and economic strength, the potency of different weapons and weapon systems, including biological ones, the integrity of cyber systems, the ability to foresee and prepare for varied scenarios.
Also read | What the world would look like in 2026 if moulded by Trump’s efforts to make America great
However, the policy prescriptions it arrived at were a bit of a letdown, as if you rubbed Aladdin’s lamp in earnest, and black smoke billowed out in huge plumes, only to reassemble in the shape of Dopey, the least prepossessing of Snow White’s seven dwarfs.
A collective Indian response to the challenge AI presents to retaining and building strategic autonomy calls for democratising society, culture and politics, besides investing in universities and research institutions.
Not that the recommendations lacked sense in themselves. Mobilise AI for scientific discovery, goes one, to link AI with research in physics, biology, medicine, and material sciences. With Google Deepmind’s Alphafold having created, in its different iterations, a database of 200-odd million forecasts of how protein molecules fold, MIT having synthesised a new antibiotic based on the database, and nuclear fusion startups using AI to calibrate the positioning and strength of magnets to control the ebb and flow of the plasma in which hydrogen atoms fuse into helium, anyone who knows anything about AI already knows that it can be used to advance science and technology.
Private sector in AI research
Another prescription is to mobilise the private sector in AI research. India spends 0.65 per cent of its GDP on R&D, which means that the total spending by the country is less than that undertaken by a single company like Amazon. Of the little that is spent on R&D, 63 per cent comes out of the public sector.
To say that India’s private sector prefers low-hanging fruit is to pay it a compliment. The Titans of India Inc actually pick up the fruit that lies on the ground, either having fallen, over-ripe, from the trees, or having passed through the alimentary canals of those who ventured to pluck and consume the lush fare on the upper terraces of the woods. Indonesia’s civet coffee makers do themselves an injustice when they fail to capitalise on the natural appetite in India for their brand of joy.
Also read | AI, Trump and the desi tech job meltdown of 2025
Scale domestic semiconductor fabs under the Semicon India Mission — that is as garbled a recommendation as you can get. India’s semiconductor mission is to create domestic manufacturing capacity, not national manufacturing capacity. Domestic manufacturing capacity is good for guarding against shortages on account of logistical disruptions. But for assured supply and beating geopolitical tech boycotts, the domestic capacity must be under national control through and through.
Domestic vs national
If Intel sets up a fab in India, that would dutifully eat up all the subsidy on offer and notch up a success for the Semicon India Mission. But if a Donald Trump-like character decides to ban India from accessing advanced American technology just because Intel produces the chips in India, it would not guarantee Indian entities access to those chips built using American technology.
For national capability in chip manufacture, India would need to indigenously produce every part of the chipmaking ecosystem, from extreme ultraviolet lithography machines of the kind produced by the Dutch company ASML, to the deposition, thermal treatment, cleaning and packaging equipment produced by Tokyo Electron and the precision optical mirrors produced by ZEISS.
Too ambitious? The Chinese have their own end-to-end chipmaking capacity now. So can India, if it puts its mind to it.
What is required is not to build domestic fabs and waste subsidies on foreign companies to locate some of their production in India, but to work out all the different components of the chipmaking ecosystem, and fund a variety of startups to come up with their own solutions to match the capability of each one of those components.
Watch/Read | Is AI psychosis worth it? Experts flag rising dependence on AI
The strategists also call for leveraging the Rs 1 trillion R&D corpus that the government announced in the 2024 budget and has fleshed out subsequently. The government has proposed to make this amount available as soft loans repayable over a long period.
Funding R&D with loans?
Funding R&D with loans is a ridiculous idea. How is someone who spends borrowed money to fund research supposed to repay the loan if the research fails? If the answer is that the government would lend this money to the equivalent of venture funds, each of which would finance a portfolio of ventures, some of which would generate enough returns to more than offset the losses on ventures that fail, that begs the question: Why should these venture funds require government funds?
Just give contributors to the National Pension System and the Employees’ Provident Fund the option of investing, say, 5 per cent of their retirement savings in one venture fund or another.
You cannot simultaneously venerate mythical flights of fancy as the literal truth of ancient Indian capability in science, complete with its cultural ecosystem of reserving knowledge as the terrain of just a privileged elite.
Yet another weakness of the recommendations is that they see AI as some standalone silo. If AI is to be effective in guiding a swarm of drones, India would need capability in not just AI but in drones as well. For AI to be effective in real-time analysis of images captured by small aperture radar on satellites in low-earth orbit, India will need extensive capability in SAR, satellites and satellite launches at scale.
Paradigm shift needed
The short point is that the use of AI to fortify national security calls for a paradigm shift in science and technology, and in culture across the board, not just in AI. You cannot simultaneously venerate mythical flights of fancy as the literal truth of ancient Indian capability in science, complete with its cultural ecosystem of reserving knowledge as the terrain of just a privileged elite.
Also read | How Aranya Sahay’s Humans in the Loop spotlights rural, invisible workers training AI
To generate a paradigm shift in science and technology, India will have to cultivate what the Constitution calls the scientific temper, and not just invest in some institutions and labs.
India’s tradition treats knowledge as finite and pre-existing, contained in the Vedas, which meant that the task before young minds was to master the texts, not to create new knowledge. This went hand in hand with acceptance of hierarchy and authority, whether inside the classroom or in social ordering.
That tradition will have to be abandoned if education is to truly become a process of cultivating critical thinking and teaching young people how to learn throughout their lives. That is the only way to bring about a paradigm shift in India’s approach to science and technology.
Ending sectarian politics
The scientific temper would put paid to the paradigm of sectarian politics, which underpins the political success of the present government in office. A collective Indian response to the challenge AI presents to retaining and building strategic autonomy calls for democratising society, culture and politics, besides investing in universities and research institutions.
Arjuna saw just the neck of the bird he was to shoot, and that was fine. He had been assigned his target. To deploy AI to strengthen national autonomy, you need to identify the target, determine the right neck of the woods, the right woods and the right season.
Two Chinese universities now lead at least one world ranking of universities. Indians are happy to glory in the historic primacy of Nalanda and Taxila. Why live in the mediocre present when the past was grand, and the future lends itself to grandiose eloquence?
(The Federal seeks to present views and opinions from all sides of the spectrum. The information, ideas or opinions in the articles are of the author and do not necessarily reflect the views of The Federal.)
