Kamal Ghorui
BENGALURU: India has emerged as the second-largest market for Claude, reflecting the country’s rapidly expanding AI ecosystem and technically sophisticated developer base. Nearly half of Claude usage in India centers on computer and mathematical tasks, including building applications, modernizing legacy systems, and deploying production-grade software—underscoring the depth of engineering engagement across the market.Anthropic has also opened its office in Bengaluru.“India represents one of the world’s most promising opportunities to bring the benefits of responsible AI to vastly more people and enterprises,” said Irina Ghose, managing director of India at Anthropic. “Already, it’s home to extraordinary technical talent, digital infrastructure at scale, and a proven track record of using technology to improve people’s lives. That’s exactly the foundation you need to make sure this technology reaches the people who can benefit from it most.”India is home to over a billion people who speak one of over a dozen officially recognized languages. However, AI models have historically demonstrated stronger performance in English than in many Indic languages. Six months ago, Anthropic launched a company-wide initiative to address this disparity by curating higher-quality, more representative training data across 10 of India’s most widely spoken languages: Hindi, Bengali, Marathi, Telugu, Tamil, Punjabi, Gujarati, Kannada, Malayalam, and Urdu.This effort has already led to measurable improvements in model fluency, with ongoing work focused on further enhancing multilingual performance.Anthropic is now collaborating with Karya and the Collective Intelligence Project to develop rigorous evaluation benchmarks that test model performance on locally relevant tasks. These evaluations span domains such as agriculture and law and are being designed in partnership with domain experts from leading Indian nonprofits, including Digital Green and Adalat AI.The evaluations will assess real-world utility in Indian contexts and are expected to inform future model improvements for Indic language speakers and sector-specific applications.
