IIT professor Mitesh Khapra featured in TIME’s 2025 most influential in AI alongside Elon Musk, Sam Altman

Before joining IIT Madras, Mitesh Khapra worked at IBM Research India, where his focus was on machine translation and deep learning.

Mitesh Khapra, Associate Professor in the Department of Computer Science and Engineering at the Indian Institute of Technology (IIT) Madras, was named among TIME magazine’s 2025 list of the 100 most influential people in artificial intelligence.

The list, considered one of the leading international recognitions in the technology sector, featured academics, entrepreneurs, policymakers, and business leaders shaping the future of AI. Khapra’s inclusion placed him alongside figures such as OpenAI chief executive Sam Altman, Meta founder Mark Zuckerberg and xAI head Elon Musk.

Unlike most other names on the list who lead multinational technology companies, Khapra was recognised primarily for his academic work. His research has focused on natural language processing and machine learning, particularly the development of open-source tools and datasets for Indian languages.

Khapra co-founded AI4Bharat, a research initiative at IIT Madras that began in 2019 with the goal of building publicly available datasets, models, and tools for India’s languages. According to TIME, “almost every Indian startup working on voice technology for regional languages uses the datasets created by Khapra and his team.”

While international AI models perform effectively in widely used languages such as Hindi and Bengali, they have often proved inadequate for less represented languages. To bridge this gap, AI4Bharat conducted one of the largest projects of its kind in India: teams visited close to 500 of the country’s 700 districts to record thousands of hours of speech across varied educational, cultural, and economic groups. This effort covered all 22 of India’s constitutionally recognised languages.

Khapra explained the shift that these datasets had brought about in Indian research. “Fifteen years ago, an average PhD student in India working on language technology would mostly focus on English-related problems,” he said. “But now, with the availability of these datasets, I see a shift, Indian students are increasingly working on challenges specific to Indian languages.”

The contributions of AI4Bharat have been widely adopted. Nearly 80 per cent of the data used in the Government of India’s Bhashini programme—a national initiative aimed at making digital services accessible in local languages—came from the group’s work.

The datasets have also been used in the Supreme Court of India for translating official documents, and in agricultural voice-based services where farmers can register subsidy concerns in their local languages. Several global technology firms have integrated AI4Bharat resources to improve performance in languages such as Hindi and Marathi.

Before joining IIT Madras, Khapra worked at IBM Research India, where his focus was on machine translation and deep learning. He earned both his MTech and PhD from IIT Bombay.

His research has been presented in leading international conferences, including the Association for Computational Linguistics (ACL), NeurIPS, and AAAI, where he has also served as an area chair. He has previously received a number of significant honours, among them the IBM PhD Fellowship, the Microsoft Rising Star Award, and the Google Faculty Research Award (2018).

In its citation, TIME highlighted that the work of Khapra and AI4Bharat had already shaped India’s digital infrastructure by ensuring that technology could be accessed in languages spoken across the country, not just in English. The magazine noted that the availability of high-quality Indian language datasets had also encouraged a generational shift in academic research priorities.