Anthropic has formally opened its Bengaluru office, marking a significant step in its India expansion strategy and underscoring the country’s growing importance in the global AI ecosystem.
India is now the second-largest market for Claude.ai, with nearly half of its usage in the country focused on computer and mathematical tasks such as building applications, modernising systems, and deploying production-grade software. The company says Indian developers represent one of the most technically intensive user bases globally.
Alongside the office launch, Anthropic announced new partnerships spanning enterprise, education, and agriculture—signalling a broader push to embed responsible AI across critical sectors.
“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 – India, Anthropic. She highlighted the country’s technical talent, digital public infrastructure, and track record of using technology for societal impact as key enablers for AI adoption at scale.
Strengthening AI for Indic languages
While over a billion people in India speak one of more than a dozen officially recognised languages, AI systems continue to perform more reliably in English. To address this, Anthropic launched a company-wide initiative six months ago to improve model performance across 10 widely spoken Indian languages, including Hindi, Bengali, Marathi, Tamil, Telugu, Punjabi, Gujarati, Kannada, Malayalam, and Urdu.
The effort focused on curating higher-quality, representative training data, leading to measurable improvements in model fluency. The company says work is ongoing.
Anthropic is now collaborating with Karya and the Collective Intelligence Project to develop evaluation frameworks that test AI performance on locally relevant tasks across domains such as agriculture and law. Domain experts from organisations including Digital Green and Adalat AI are contributing to the initiative.
The evaluations are expected to guide future model improvements for Indic-language speakers and India-specific use cases. Anthropic also intends to make these benchmarks publicly available.