If the AI race were a Bollywood potboiler, India ended up with a ‘happily-ever-after’ with February’s India AI Impact Summit jamboree in Delhi.
Some embarrassments apart, the event was a success—lakhs of people thronged venues like the Bharat Mandapam, there was breathless media and public interest from around the world, and there was a line up of dignitaries and tech tycoons ranging from French President Emmanuel Macron to ChatGPT’s Sam Altman flanking Prime Minister Narendra Modi on the dais. It seemed like India had arrived on the global AI arena. The New Delhi Declaration at the end of the summit saw 91 nations and two international organisations signing up to work for an inclusive, secure and sustainable AI ecosystem.
Unfortunately, the AI race is more like a long-running web series than a quick movie climax. And if the initial seasons are anything to go by, the world’s most populous country is only playing a supporting role as a big consumer market, and still needs a plot twist to rise up to a starring role.
That is a hard knock for a nation that had gotten used to being the top dog in the tech pack, mainly as an IT services powerhouse. The sudden turn in technology towards AI has exposed India’s Achilles heel: its sore lack of skills, scale, capability and know-how, which could hurt its power and position en route to Viksit Bharat 2047.
The alarm bells have been ringing for a while in the corridors of power, with Modi himself making it clear during the AI Impact Summit: “We are approaching this not as a future problem but as a present imperative…. AI presents both a tremendous opportunity and a challenge for the IT sector. My vision is that India should be among the top three AI superpowers globally, not just in consumption, but in creation.”
He has a point. There was much fanfare when an Ernst & Young report put India’s AI adoption score at 53, way higher than the global average of 34, or when Stanford University recently placed India at no. 3, just after the US and China, in what it mysteriously calls an AI ‘vibrancy’ index. But the reality is a wake-up call: Indians are using a lot of AI, but are not creating anything cutting-edge in the space.
ChatGPT’s highest user base outside the US is in India, at more than 10 crore. This probably weighed on the mind of Sam Altman, CEO of its parent company OpenAI. After having billed India’s chances in the AI race as ‘hopeless’, he changed his stance just before his appearance at the summit, saying that India had the potential to be a ‘full stack AI leader’.
The odds are stacked against it right now. For one, the dominance of foreign players (US and China) is too significant, considering the heavy entry barrier ranging from computing power and requirement of resources—from electricity to water—that data centres and chip plants are thirsty for. The current geostrategic scenario means that India has access to just 38,000 high-performing graphic processing units (GPUs) that are central to the computing required for AI and large language models (LLM). Information Technology Minister Ashwini Vaishnaw announced at the summit that India will get access to an additional 20,000 GPUs in the coming weeks.
Then there is the question of skills and the knowledge base itself. Despite the plethora of ads for AI skills and crash courses, experts estimate that India might only have a few hundred qualified AI engineers. India spends a paltry 0.7 per cent of its GDP on research and development, compared to AI rivals like South Korea (above 5 per cent of GDP) and Japan (around 3.5 per cent).
“A lot of us are concerned about how India will fare in the AI era,” said India’s chief economic adviser V. Anantha Nageswaran. “The issues are… data centre resource consumption, the kind of financial requirement for the LLMs in the current ecosystem of the western world, greater localisation for domestic value creation, of course the macro question of mitigating labour market impact, navigating geopolitical constraints to hardware access including chips and, importantly, for geopolitical and security perspectives, can we continue depending on foreign models in public administration and critical sectors.”
The stakes go way beyond just getting a share of AI’s technological rewards. Perhaps like oil in the last century and the internet in the early years of this one, AI is poised to be the building block of the next wave of development, and that means the cost of losing this race would be existential. As Srinivas Padmanabhuni, CTO of AIEnsured, an AI testing platform, said, “In today’s geopolitical context, leading in AI means shaping the laws and ethics of the future. To reach the top, India must strengthen key areas: build deep AI academic capacity with large-scale cloud infrastructure, position itself as the AI voice of the Global South; create a multi-billion-dollar risk capital pool to back high-impact startups, and launch a national AI skill mission.”
How is that possible? The way out, said Kalyan Sivalenka, MD at Hyderabad Angels Fund, is to not try to outspend the US or China, but out-specialise them by using resources effectively. “We need deeper public-private partnerships to subsidise compute access for startups. The ‘AI Mission’ must move from policy to physical hardware on the ground,” he said. “We need to see more domestic family offices and tier-1 venture capitalists moving away from ‘fast-exit’ mindsets and toward ‘deep-tech’ bets. The government’s recent announcement of Rs1 lakh crore for deep tech startups is a good move.”
The government is on it. The IndiaAI Mission is at the core of the efforts to develop indigenous models and spur a domestic innovation boom, with budgets over the last couple of years allocating more and more for this space. In fact, this year’s Economic Survey mentioned AI nearly 400 times, saying, “There is no time to waste.”
Calling AI a core economic strategy instead of being just a technological race, the Economic Survey also suggested a way for India to jump in and gain a foothold on the AI gravy train: forgo the western model of top-down heavy investment in AI models, and instead focus on specific models that run on frugal expenses using local hardware and are aimed at solving uniquely Indian problems, ranging from agriculture to health care, and more. It is a theme that popped up over and over at the AI Summit as well.
“AI demands responsible adoption: it is critical for India to build a governance framework for the benefit of people, policy, and most importantly, country. Else it could derail and destroy both value and ethics,” warned Padmaja Ruparel, co-founder of IAN Group, India’s largest angel investor network.
The potential is there, with enterprise use cases showing a steady trickle that could turn into a torrent of innovation in the coming months and years. Take, for example, multinational biggie Dassault Systemes’s ‘Virtual Twins’ model that is already at work in India in firms like L&T. “Virtual Twin is a live digital version of a product, plant or project that learns from real‑world data,” explained Deepak N.G., MD, Dassault Systemes India. “AI looks at patterns in that data and gives practical suggestions like ‘this part may fail soon’, ‘you will need more of this material next week’, or ‘send this truck on a different route’, so Indian businesses can design better, run smoother and take smarter decisions every day.”
India’s challenges are fundamentally two-fold: manoeuvre within the constraints of funding and technology to come up with innovations that will take it to the next level in the AI sweepstakes, even while dealing nimbly with the question of job losses and re-skilling. “It is imperative that citizens are upskilled for AI, else we will see job losses. Transitioning into new technologies is required, and this programme will need to be scaled up across the country,” said Ruparel.
Said Vinay Maheshwari, an entrepreneur, investor and business transformation strategist: “India’s trajectory is promising, but structural gaps need urgent correction. The next 18 to 24 months will determine whether India remains a large AI market or [transforms into] a global AI innovation leader.”