As AI technology diffuses across the world, India is seemingly making good use of its groundbreaking capabilities .
Frontier AI firm OpenAI, in a recent report, highlighted that Indian users are among top adopters of ChatGPT’s advanced capabilities including data analysis, coding and education.
Higher Learning
Indian users have closed much of what it defines as ‘capability overhang’, the gap between what AI is capable of doing and how much of the capabilities are being utilised by the users, it said.
In fact, Indian users are nearly 3x above the global median when it comes to asking ChatGPT questions about coding and nearly 2x above when it comes to questions about education and learning. They are also slightly higher than the global average when it comes to work related queries using the model.
OpenAI, in its report, also flagged that AI adoption in India is three times more concentrated in cities than other countries. This indicates that while India has competitive AI capability, its deepest advantages are clustering in a small number of urban hubs, including Bengaluru, Hyderabad, Delhi, and Chennai.
Enterprise Tech
It also added that India is one among the strongest users of its coding tool Codex, with the app having seen a 4x growth in users in two months.
Speaking to businessline, analysts and industry experts believe that beyond a large base of engineering talent, the advanced AI usage in India is driven by a combination of factors including limitations in access to specialised experts, tools or institutional support and the cost sensitivity of Indian users trying to maximise value out of each interaction.
Sandeep Chordia, Chief Operating Officer at Kotak Securities says an ever increasing number of digitally native businesses mean that people are comfortable experimenting with next gen tech like AI. “This naturally drives deeper usage of AI for coding, analytics, business problem-solving, content creation and productivity. More importantly, the use cases are more outcome-oriented with less focus on experimenting but actively applying AI to solve problems,” he said.
Certain missing layers within India’s enterprise tech stack also contribute to advanced use of AI.
Self-taught
As Biswajeet Mahapatra, Principal Analyst, Forrester puts it, global organisations often have a ‘readymade’ expert layer like pre-configured co-pilots or formal workforce training unlike Indian employees who have to learn it by themselves..
“Because these support systems are often absent or still being established, users in India frequently rely on reasoning-heavy interactions. This involves building their own multi-step prompt sequences leading to more iterative and complex reasoning,” he said.
The value-conscious nature of Indian users also mean that they push more detailed prompts and expect comprehensive responses in fewer attempts.
“Indian users tend to extract more value per interaction by packing more context and multiple requests into a single prompt. That naturally increases complexity and reasoning usage, even if the number of queries itself is not higher,” Amit Khanna, Partner and Lead - dGTL and AI, Grant Thornton Bharat
Richer Context
The linguistic gap between the use case and the model is also a major contributor.
Mahapatra believes that most incumbent LLMs often see a drop in performance in languages with minimal training data - like many Indian languages.
“To then achieve accurate results, Indian users frequently provide richer contextual cues within their prompts to ensure the model captures specific cultural nuances. This requirement for ‘contextualisation buckets in non-English or localised contexts significantly inflates both the number and type of tokens used,” he said.
Published on April 22, 2026