AI adoption in India may impact 38 million jobs: report

Generative artificial intelligence (GenAI) is set to revolutionize India’s workforce and reshape its economic landscape by 2030. AI adoption could impact 38 million jobs, contributing a 2.61% productivity boost to the economy through the organized sector, with an additional 2.82% potential growth from GenAI adoption in the unorganized sector.

According to the latest Ernst & Young (EY) India report, intelligence costs have been reduced, thanks to the open-source movement and the trend to use purpose-specific small language models (SLMs). 

This is making AI accessible to smaller businesses, and very soon, it may become possible to use open-source models for as low as a few thousand rupees a month in India, the report said. 

However, enterprise adoption of GenAI remains very low. According to the EY survey, 36% of Indian enterprises have allocated budgets and started investing in GenAI, while another 24% are exploring its potential. The technology sector is leading the adoption, with life sciences and financial services following. However, the business value is still limited, as only 15% have GenAI workloads in production, and just 8% can fully measure and allocate AI costs.

“GenAI is transforming India’s economic landscape by unlocking unprecedented opportunities across sectors. This revolution will fundamentally reshape jobs, driving productivity and innovation. Building talent pipelines and prioritizing upskilling must be at the forefront of every organisation. By fostering public-private collaborations and investing in talent development, India can also become a global hub for AI skilled talent,” Rajiv Memani, chairman and chief executive, EY India, said in a statement. 

GenAI is poised to transform every job, driving substantial productivity and economic growth. The EY India report indicates that 24% of tasks across industries could be fully automated, while time spent on 42% could be significantly reduced with AI, freeing up 8-10 hours per week for corporate workers. This translates to a productivity boost of 2.61% by 2030 in the organized sector, affecting 38 million Indian employees and an additional 2.82% in the unorganized sector.

At the industry level, the services sector is expected to see the biggest productivity boosts due to its higher labor share of gross output, while manufacturing and construction will experience more modest gains. Even in these sectors, AI can enhance efficiency by improving capital deployment and resource utilization, ultimately reducing labor costs and boosting overall cost-effectiveness. To unlock this potential, businesses must rethink processes and redefine and reskill the workforce. 

According to EY, successful GenAI adoption requires clear strategies, pilot use cases, scalable solutions, reimagining processes, key performance index, and focused reskilling. Large-scale upskilling initiatives, supported by public-private partnerships and AI-focused training programs, are essential to closing the skill gap. With investments in skills, data, and infrastructure, GenAI can drive economic productivity and prepare India’s workforce for the future, EY said in its report. 

Skill shortage as major obstacle

Based on a survey of over 125 C-suite executives, the EY India findings highlight that skill shortages are a major obstacle to AI adoption. Only 3% of Indian enterprises have the in-house talent and resources to fully leverage AI, while 97% of executives identify talent gaps as a key challenge.

“In industries like financial services, healthcare and retail, AI will reshape basic processes including customer acquisition, operations and service, while IT/ITeS and BPO will undergo more dramatic changes. Next-generation industries like biotech, advanced manufacturing and renewables will have the potential to leapfrog to AI-first business models. To maximise the potential for economic growth, India needs to focus on AI policy agenda, compute infrastructure, AI research, addressing challenges in responsible governance, intellectual property rights, and data protection,” said Mahesh Makhija, Technology Consulting Leader in EY India.

EY’s analysis of over 10,000 tasks across industries reveals varying productivity gains. Call center management is expected to see the largest impact, with an 80% productivity boost, followed by software development at 61%. Content development and distribution will grow by 45%, customer services by 44%, and sales and marketing by 41%. IT/ITeS is projected to experience a 19% productivity increase, healthcare 13%, and banking/insurance 8-9%. Auto and pharma industries are expected to see a modest 2% boost due to their lower labor contribution to gross output.

AI costs to continue decreasing

The cost of using AI models has already plummeted, making them increasingly accessible to enterprises. According to EY, OpenAI’s GPT API costs, for example, have dropped nearly 80% in two years, while open-source releases like Meta’s (NASDAQ: META) Llama are unlocking new capabilities. 

The EY report said that this cost is expected to fall to around Rs 120 per hour ($1.39) or lower as India-specific large language model (LLM) offerings become viable.

India has already seen a surge in Indic LLMs that use open-source models fine-tuned with Indian language datasets. A notable initiative is Bhashini, an Indian federal government-led AI project focused on creating an open-source Indic language dataset to improve the Internet and digital service accessibility in Indian languages.

Looking ahead, AI is set to become a key component of the India Stack, serving as digital public infrastructure to support next-generation platforms. According to the EY report, the growing GenAI startup ecosystem and local AI infrastructure will drive adoption among Indian enterprises.

Recommendations

“The India Dataset Platform should be launched as soon as possible, even in a beta version, to enable early feedback from the AI ecosystem for continuous improvement,” the EY India report stated. 

“The platform can be modelled after successful examples like Hugging Face, with datasets tailored to key sectors such as payments, healthcare, education, and agriculture, readily available for AI development,” the report said.

The platform should adopt a privacy-first approach, enabling crowdsourcing for AI data needs, fostering ecosystem collaboration (similar to ImageNet), and serving as a marketplace for private datasets.

The government should expedite the deployment of AI use-case development, particularly in sectors like healthcare, agriculture, and education.

“India has initiated the procurement of GPUs. Given the growing demands in the AI ecosystem, it is crucial to fast-track this process,” the EY report recommended.

IndiaAI, a federal government initiative launched in March 2024, has selected ten firms to procure 10,000 graphics processing units (GPUs) for the final bidding process. 

The IndiaAI Mission aims to procure over 10,000 GPUs in a public-private partnership to address India’s computing capacity needs and develop a high-end AI ecosystem. GPUs are the backbone of AI processing and are becoming increasingly vital in data centers. Offering top-tier performance for AI training and inference, GPUs are fueling companies’ demand and pushing them to invest in expanded storage and computing capacities.

In March 2024, India approved roughly $1.24 billion for the IndiaAI Mission. The initiative has earmarked 44% of the budget to build a compute capacity of over 10,000 GPUs over five years, to strengthen AI infrastructure and accelerate advancements in AI technology nationwide.

In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership—allowing it to keep data safe while also guaranteeing the immutability of data. Check out CoinGeek’s coverage on this emerging tech to learn more why Enterprise blockchain will be the backbone of AI.

Watch: Blockchain, IPv6, AI & 5G will pave the way for the new Internet

title=”YouTube video player” frameborder=”0″ allow=”accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share” referrerpolicy=”strict-origin-when-cross-origin” allowfullscreen>