India had the world’s second-largest pool of top AI authors and inventors in 2025, trailing only the United States, according to a new study by Stanford University.The AI Index 2026 report published on Monday, April 13, by the Stanford Institute for Human-Centered Artificial Intelligence (HAI) found that the largest share of identified AI authors and inventors came from the United States (220,520) last year, followed by India (50,460) and Germany (48,520).While the US is home to the most number of AI researchers, the flow of these experts into the country is dramatically slowing. “The number of AI scholars moving to the United States has dropped 89 per cent since 2017 […] down 80 per cent in the last year alone,” as per the annual report.Meanwhile, India saw the largest net outflows (-16.9) of AI research talent in 2025. This metric reflects the difference between the number of AI authors and inventors who move to or out of their respective countries. “The near mirror-image relationship between the two countries reflects the well-documented fact that the US has been the main destination for Indian AI talent,” the report said.These figures come at a time of intensifying global competition in AI, where India is looking to secure a stronger foothold. They take on added significance in the context of other factors, including immigration policy and geographic distribution of investment and employment.Top AI authors and inventors in countries such as India and Brazil tend to have fewer PhD holders and a more mixed range of educational backgrounds, as per the report. There is also a gender gap among AI authors and inventors evident across all countries, with men making up the majority in all cases, though the extent of the gap varies by country.Other India-specific findingsThe enterprise adoption of AI continued to grow in 2025, with organisations in developing markets such as India reporting an 88 per cent adoption rate last year, up from 77 per cent in 2024. “In 2025, 58 percent of employees globally reported using AI at work on a semiregular or regular basis, but in India, China, Nigeria, the United Arab Emirates, and Saudi Arabia, the share exceeded 80 per cent,” the report noted.Story continues below this ad (Image: Stanford University report)On AI skilling of the workforce, India saw the highest skill penetration rate (3.0) followed by the United States at 2.0 and Germany at 1.8. This means that AI skills were listed in LinkedIn profiles of Indian users at almost three times the global average. However, these countries “showed a persistent gender gap when measuring male and female AI skill penetration rates against the global average.” For instance, men in India list AI skills at more than 1.5 times the rate of women, as per the report.While AI enterprise adoption and AI skill penetration rates were high in India last year, the country also saw the sharpest rise in AI nervousness of any country surveyed. This signals that a growing share of AI users in India reported feeling anxious about using these products and services, rather than emphasising their benefits or drawbacks.Also Read | Top AI chatbots endorsed harmful user behaviour in 51% of cases, new Stanford study finds“Between 2024 and 2025, India registered the sharpest rise in concern around AI usage (+14 percentage points) with only a modest increase in excitement (+2),” the report said. Globally, the share of respondents who say AI products and services offer more benefits than drawbacks rose from 55 per cent in 2024 to 59 per cent in 2025, and the share saying these products make them nervous increased to 52 per cent, as per the report. (Image: Stanford University report)India attracted a total of $4.09 billion in global private investments in AI last year, trailing countries such as the US ($285 billion), China ($12.41 billion), United Kingdom ($5.90 billion), France ($4.36 billion), and Canada ($4.28 billion). India also saw 108 newly funded AI companies in 2025, with the US once again leading by a wide margin at 1,953.Story continues below this adKey global findings– Open-source AI: Over 90 per cent of the most ‘notable’ AI models in 2025 were produced by tech companies, but the most capable models are also the “least transparent,” the report noted. It acknowledges that open-source AI development continues to scale with uploads on platforms such as GitHub and Hugging Face tripling to 5.6 million projects since 2023. Out of these, India represented 5.2 per cent of open-source AI projects that have secured at least 10 stars on GitHub.– Model parameters, context windows: The parameters of AI models have idled near the 1 trillion-mark for the past three years, though this might be because frontier AI labs are sharing less information. Model context windows, on the other hand, have grown almost 30x per year since mid-2023. This means that AI models that once accepted a few thousand tokens can now process millions or more. (Image: Stanford University report)– Global AI compute capacity: Global AI compute capacity has grown 3.3 times per year since 2022, reaching the equivalent of 17.1 million H100 chips with Nvidia accounting for over 60 per cent of total compute today. Google and Amazon chips make up the remainder of the supply, and Huawei holds a small but growing share of the chip market, as per the report.Also Read | India ranks third in Stanford University’s 2025 Global AI Vibrancy tool, this country is number 1– Global data centre capacity: AI data centres are concentrated in select regions, with the US hosting 5,427 data centres, more than ten times any other country, and consuming more energy than any other region, up till 2025.Story continues below this ad– Global AI investment: Private investments more than doubled in 2025, with generative AI accounting for 60 per cent of all private AI funding. Newly funded AI companies rose 71 per cent, and billion-dollar funding events also doubled. US investment exceeded the combined total of China and Europe by a wide margin. However, US public-sector investments were a modest $20.4 billion in AI-related contracts and grants, against $285.9 billion in US private investment in 2025 alone. (Image: Stanford University report)– AI benchmarking: The report highlighted that top benchmarks used to measure AI progress saw error rates up to 42 per cent.– Barriers to AI adoption: Gaps in knowledge (59 per cent), budget constraints (48 per cent), and regulatory uncertainty (41 per cent) were identified as the main obstacles to the adoption of AI at workplaces. However, the report also noted that AI-specific governance roles grew 17 per cent in 2025, and the share of businesses with no responsible AI policies in place fell sharply from 24 per cent to 11 per cent.