India’s IT services industry is likely to undergo a significant workforce transition because of artificial intelligence (AI), but fears of widespread job losses are premature, according to Neelkanth Mishra, member, Economic Advisory Council to the Prime Minister (EAC-PM), and executive director-designate, World Bank.
Mishra said the industry should be viewed holistically by combining traditional IT services companies with Global Capability Centres (GCCs). He noted that India’s services exports grew 15 per cent in dollar terms in April and have been expanding faster than the country’s gross domestic product (GDP).
Mishra was speaking to the media at the launch of the Internet and Mobile Association of India (IAMAI)’s AI Council of India in Mumbai.
“The organisational boundary itself may be shifting because of AI,” Mishra said, explaining that some work previously outsourced to IT services firms is increasingly being brought in-house. As a result, he said, the industry’s performance should be assessed by looking at both IT services firms and GCCs together.
He acknowledged that AI poses significant transition challenges for the sector, particularly around reskilling. Companies will need fewer people focused purely on coding and more employees skilled in areas such as design, user interface (UI) and user experience (UX), he said.
Mishra also said most layoffs so far have been concentrated in software companies because labour is their largest cost and software development is increasingly being automated. However, he argued that lower software development costs could also increase the volume of software being built, invoking the “Jevons paradox”, where lower costs drive higher consumption.
“This is not to say there will be no disruption. There is a lot of restructuring and strategic shifts that firms need to undertake. But the current panic and paranoia about massive job losses are, in my view, still premature,” he said.
In a fireside chat with Sharad Sanghi, founder and chief executive officer (CEO) of Neysa, Mishra also addressed the constraints and opportunities for making India a leading country in AI.
India must prioritise building a merit-based research ecosystem and deepen risk capital if it wants to emerge as a global leader in AI, while simultaneously developing domestic capabilities across the AI value chain, Mishra said.
Mishra said India currently lacks sufficient cutting-edge research and described a merit-based research ecosystem as the single most important requirement over the next three to five years if the country wants to move beyond being merely a consumer of AI technologies.
Mishra was speaking in a fireside chat with Neysa founder and CEO Sharad Sanghi at the launch of the AI Council of India in Mumbai, organised by IAMAI.
Mishra said capital and talent remain India’s biggest constraints, with capital being the more immediate challenge. Drawing from his experience on the government’s fund manager selection committee under the Fund of Funds, Mishra said deploying capital into deep-technology ventures remains a challenge because India has too few investors with the expertise to evaluate semiconductor and AI start-ups.
“There are good fund managers, but not enough to absorb the amount of capital available. We have funds that have managed Rs 100 crore or Rs 150 crore asking for Rs 500 crore. Even if they plan to raise another Rs 500 crore on their own, they are still making a significant leap,” he said when asked about the need for higher risk capital.
He pointed to another issue: domain expertise. “Deep technology requires specialised knowledge. Most venture capital firms in India have built their track records investing in e-commerce, fintech and enterprise software. Assessing semiconductor companies or AI infrastructure businesses requires a very different understanding,” he added.
The finance minister announced this Fund of Funds in the Budget on February 1, 2024. It took about one and a half years to operationalise it. “Another seven or eight months were needed to complete the institutional framework, and this has actually been one of the fastest government implementations I have seen. Now we have reached the stage where we are ready to deploy capital. The challenge is finding enough qualified fund managers,” he said.
According to Mishra, India must also encourage experienced Indian-origin professionals working overseas to return and build technology companies at home while creating many more deep-tech start-ups through universities and research institutions.
Beyond capital, Mishra said India must build capabilities across the AI value chain instead of focusing only on foundation models or data sovereignty.
He cited the recent restrictions around Anthropic’s Mythos as an example. “The recent restrictions around Anthropic’s Mythos models were eventually lifted because the company argued that the security loopholes identified were relatively minor, whereas the US government viewed them differently. In that case, the restrictions were primarily about safety. But what if, in the future, such restrictions are imposed for strategic reasons?”
Mishra added that this is precisely why ownership matters. “We need companies where the intellectual property is owned and controlled from India. When we talk about India’s leadership in AI, the opportunity is not just to consume AI but to build products for India and for the world,” he said.
He argued that AI should be viewed as a strategic capability spanning chip design, electronics manufacturing, data centres, energy infrastructure and applications, particularly as global access to advanced technologies could become increasingly restricted.
“From a national security perspective, it is important for India to have its own models. It is equally important for us to have a presence across the AI value chain,” he said.
Using semiconductors as an example, Mishra said while manufacturing facilities are important for supply chain resilience, much of the economic value accrues from chip design. Similarly, AI sovereignty extends beyond keeping data within the country and includes critical infrastructure such as power, cooling systems and affordable energy required to support compute-intensive workloads.
He also highlighted the role of the India Semiconductor Mission (ISM), saying partnerships with global semiconductor companies would be essential if India wants to compress the development cycle. China had begun building its semiconductor ecosystem nearly 25 years before India, he noted, and the country cannot afford to take another quarter century to catch up.
Mishra said ISM is working to strengthen India’s semiconductor design ecosystem, expand incentives under the Design Linked Incentive scheme and encourage experienced professionals overseas to return. He also called for greater awareness of these initiatives and more specialised venture capital to back semiconductor start-ups.
“We need partners. We need to work with global semiconductor leaders and identify what we can do to encourage them to establish operations here,” he said, adding that while India should leverage global technologies, it must simultaneously develop domestic capabilities to reduce strategic dependence over the long term.
Box:
The AI Council of India (AICI), a national platform and a council of the Internet and Mobile Association of India (IAMAI), was launched in the presence of Ashish Shelar, Maharashtra’s minister for information technology and cultural affairs, and Neelkanth Mishra, member, Economic Advisory Council to the Prime Minister, and executive director-designate, World Bank. The council will create synergy across India’s rapidly evolving artificial intelligence ecosystem.
AICI brings together policymakers, technology companies, AI start-ups, academia, investors and other key stakeholders on a common platform to foster collaboration, accelerate AI adoption and strengthen India’s leadership in artificial intelligence.
The council’s work will be anchored around three strategic focus areas. First, advancing models, research and frontier AI, including foundation models, sovereign AI development and localised benchmarks for India’s more than 500 million non-English users. Second, strengthening AI infrastructure and compute by expanding access to GPU clusters and reducing compute costs to make AI development viable for start-ups and MSMEs alike. Third, driving applied AI and lean innovation by scaling cost-effective, open source-led AI solutions across healthcare, agriculture, manufacturing, financial services and education, sectors where India’s AI adoption significantly lags global benchmarks.
