Anthropic Bengaluru Office Anchors India Expansion

The opening of the Anthropic Bengaluru office marks a major step in the company’s India strategy. The new facility positions India as its second-largest market globally and its second location in Asia after Tokyo.

India’s developer base is central to this expansion. Nearly half of Claude.ai usage in the country is focused on complex computing and mathematical tasks. These include building applications, modernising systems and shipping production software. The usage pattern reflects a technically mature user base rather than casual experimentation.

Irina Ghose, MD, India, Anthropic, said India offers the scale, technical talent and digital infrastructure necessary to extend responsible AI adoption across enterprises and individuals.

Building multilingual AI for scale

India’s linguistic diversity remains a structural challenge for AI deployment. While English dominates most AI systems, more than a billion people speak one of over a dozen officially recognised Indian languages.

Over the past six months, Anthropic has improved Claude’s fluency in ten widely spoken Indian languages:

• Hindi

• Bengali

• Marathi

• Telugu

• Tamil

• Punjabi

• Gujarati

• Kannada

• Malayalam

• Urdu

In partnership with Karya and the Collective Intelligence Project, the company is developing evaluation benchmarks tailored to Indian use cases. These tests will measure model performance in sectors such as agriculture and law.

Nonprofits including Digital Green and Adalat AI are contributing to these assessments. The evaluation results will be released publicly to guide future model improvements and contribute to broader AI research.

Enterprise and startup momentum

Since announcing its India expansion in October 2025, Anthropic’s run-rate revenue in the country has doubled. Adoption spans enterprises, digital-native firms and startups.

Enterprise deployments include:

• Air India using Claude Code to accelerate custom software development and improve operational efficiency.

• CRED reporting twice the feature delivery speed and 10 percent improvement in test coverage.

• Cognizant deploying Claude across 350,000 employees globally to modernise legacy systems.

Startup integrations reflect operational use rather than pilot experiments:

• Razorpay embedding AI into risk and operational systems.

• Rocket enabling non-technical teams to build production-ready applications in hours.

• Enterpret integrating AI assistants into customer insight workflows.

• Emergent reaching $25 million in ARR and two million users within five months, building its platform entirely with Claude.

The pattern suggests that Claude is being used as a development and integration layer inside production environments.

Expanding AI access in education

Educational tasks account for 12 percent of Claude.ai usage in India. Anthropic is working with Pratham to power the Anytime Testing Machine, piloted across 20 schools covering 1,500 students. The programme is expected to expand to 100 schools by the end of 2026.

The company is also collaborating with the Central Square Foundation to provide technical mentorship and API credits to EdTech organisations serving underserved students.

The focus is practical deployment rather than research demonstration. Tools are being embedded directly into classroom environments.

Public sector and social impact integration

Anthropic is working with the EkStep Foundation to explore AI applications in agriculture, a sector employing nearly half of India’s workforce.

Additional collaborations include:

• Noora Health and Intelehealth to improve healthcare access in remote communities.

• Adalat AI launching a national WhatsApp helpline providing case updates, document summarisation and interactive legal document querying in native languages.

These deployments reflect AI use in frontline services rather than internal automation alone.

Open standards and interoperability

Anthropic developed the Model Context Protocol as an open-source standard to connect AI systems with external tools. India has become an early public-sector adopter.

The Ministry of Statistics and Programme Implementation, supported by Bharat Digital, launched the first government MCP server. The server provides open access to authoritative national statistics.

Private-sector adoption includes Swiggy, enabling grocery ordering and reservations directly via Claude.

The move indicates growing interest in interoperability standards that allow AI systems to interact securely with external databases and services.

A strategic India hub

The Anthropic Bengaluru office, led by Irina Ghose, will serve as a centre for enterprise engagement, startup partnerships, education initiatives and public-sector collaborations.

The expansion underscores India’s role as a strategic AI development hub. The scale of developer usage, multilingual research and institutional adoption signals a shift from exploratory pilots to structured, production-level integration.

With enterprise revenue doubling, education pilots expanding and public-sector standards taking shape, the Anthropic Bengaluru office represents more than a regional branch. It marks India’s growing influence in shaping applied and responsible AI deployment at scale.