Anthropic Establishes First India Office in Bangalore
US-based AI firm Anthropic has opened its first office in India, located in Bengaluru. The move follows a recent $30 billion fundraise that valued the company at $380 billion and aims to scale its enterprise and developer operations in what it identifies as one of its fastest-growing markets.
- The company was founded in 2021 by seven former senior members of OpenAI, including siblings Dario Amodei (former VP of Research at OpenAI) and Daniela Amodei. They left due to concerns about the direction of AI development, wanting to prioritize a safety-first approach. - The new Bangalore office is Anthropic's second in Asia (after Tokyo) and will be led by Irina Ghose, who previously served as Managing Director of Microsoft India, where she focused on enterprise AI adoption. - India is the second-largest market for Anthropic's AI assistant, Claude.ai, and the company's run-rate revenue in the country has doubled since October 2025. Nearly half of Claude's usage in India is for computer and mathematical tasks like building applications and modernizing systems. - The recent $30 billion funding round was led by investors including GIC, Coatue, D. E. Shaw Ventures, and ICONIQ. Anthropic also has substantial backing from major tech companies, including Amazon ($8 billion), Google ($2 billion), Microsoft, and Nvidia. - Anthropic has a strong focus on developer tools, with its Claude Code product generating a run-rate revenue of $2.5 billion. Indian companies like Air India are using Claude Code for faster software development, and CRED has used it to achieve two times faster feature delivery. - As part of its India strategy, the company is actively working to improve its models' performance in 10 widely spoken Indian languages, including Hindi, Bengali, Tamil, and Kannada. - The company is collaborating with the Indian public sector, including the Ministry of Statistics and Programme Implementation, which is using Anthropic's technology to build a server for querying economic data. - A core part of the founders' technical approach is "Constitutional AI," a method for training models to be helpful, harmless, and honest, guided by a set of principles to align with human values.
