“Solutions in India can be templates for Global South,” says Irina Ghose, MD, Anthropic India

As Artificial Intelligence (AI) moves from experimentation to enterprise-wide adoption, businesses must look beyond increasingly capable models to address trust, safety, data governance and measurable value. These are particularly significant in India given its vast technology ecosystem, diverse needs and potential to develop AI solutions at population scale.

For Anthropic, the company behind Claude, which identifies India as its second-largest market, building a presence goes beyond market expansion. It involves helping enterprises modernise, partnering with Indian IT services firms to develop AI-powered solutions, supporting developers and exploring applications across sectors such as education, healthcare and life sciences. Local inference capabilities are also expected to address data processing and governance needs, particularly in regulated industries.

Irina Ghose, who is spearheading Anthropic’s India business, has moved into the faster-moving environment of a global AI company after more than two decades at Microsoft. As she speaks about Anthropic’s India ambitions, she appears visibly excited. Her leadership in steering Anthropic’s India expansion has also earned her recognition as a Super Achiever in Business Today’s Most Powerful Women list, underscoring her growing influence in shaping India’s AI landscape.

In this conversation with Business Today, she discusses Anthropic’s India strategy, the role of partnerships in accelerating enterprise adoption, the importance of responsible AI development and opportunities for India to shape how AI is built and used globally. Edited excerpts.

 

Q: How has been the journey of building Anthropic India from scratch after coming from an established business like Microsoft?

It has been very interesting. At Microsoft, where I spent more than two decades, I was involved in early stages of shipping as we moved into cloud, established AI initiatives and built the business around them. At Anthropic, a global start-up, the pace, mindset and everything else are very different.

First, there is no playbook. You might think you can replicate what worked in other organisations, but here you have to write the playbook yourself and edit it frequently.

Second, you need to stay closely connected with teams building everything in San Francisco and people using it in India. It is a two-way signal: translating the product for users, gathering feedback on how India uses it and taking that feedback back to the teams.

You also need curiosity and a learner’s mindset. In the last two weeks, we launched three models. In a large organisation, you might wait for enablement and support to be ready. Here, you need to keep learning.

Finally, you must be a builder and be scrappy. You need an end goal, a compass and navigation, while remaining ready to change routes and stay flexible. These are the changes I have adapted to at Anthropic, and I am thriving here.

 

Q: AI is moving very fast, and there has been discussion about slowing down. What is your view, and how are you keeping pace with the development?

Dario’s essay, Pacing the Frontier, outlines the ethos of ensuring that whatever we build is safe, reliable and trustworthy. The entire ecosystem of model providers, not just one organisation, must embed these principles in its work.

Countries and democracies must also work alongside organisations and think carefully about what they put into people’s hands. Anthropic has built mechanisms to regulate and monitor what we do and keep everyone honest as we build together.

Anthropic is focused on enterprises, which value trust above all else because their brands depend on it. Protecting trust, reliability and security must take priority. These principles must be reflected in what we put into customers’ hands, and we must build together with them.

 

Q: India is Claude’s second-largest market. What have you achieved in India over the last nine to ten months, particularly in enterprise adoption?

India was chosen because it deeply aligns with Anthropic’s mission. As a public-benefit company focused on global good, Anthropic aligns with India’s ambition to drive nuanced use cases for the last mile.

I want to focus on three areas: the enterprise journey, ecosystem creation, and working with the government in priority sectors.

For enterprises across financial services, government and healthcare, key concerns include where data resides, how it is processed and whether it leaves the country during inferencing. We worked closely with our San Francisco teams and AWS on engineering required to bring inferencing to India. We announced this in August and went live in September. This will unlock opportunities in regulated segments, including banking, manufacturing, telecom and government.

Second, we are building the ecosystem by bringing in applied AI team members, customer success managers and people focused on policy and beneficial deployment. They will work with partners, venture capitalists and funds to scale our efforts.

Third, we are working with the government on beneficial deployments across education, life sciences and healthcare.

We believe replicable solutions developed in India for population-scale problems can serve the country and become templates for the Global South.

 

Q: With major investments from Google and other companies, do you think the road ahead will be challenging?

We need to look at AI’s evolution across three areas: ensuring relevance to the country, solving complex frontier use cases, and building partner ecosystems that serve Indian customers while creating templates for global markets.

For relevance and last-mile usage, we are working with Karya and others to make Claude available in more than 10 languages.

For enterprises, Claude excels at genuinely complex use cases. With inferencing available in India, we are working with banks and manufacturers on modernisation and agentic applications that improve efficiency and support people working alongside AI.

The third area is partnerships. India has a large global systems integrator (GSI), advisory and solution-partner ecosystem.

Q: How will Anthropic’s partnerships with TCS, Infosys, HCL and other Indian IT services companies influence AI adoption?

We partner with major global systems integrators in India, including TCS, Infosys, HCL, Cognizant, UST, LTM and LTTS. These partnerships have three elements.

First, partners adopt the technology themselves as customer zero and build skills through certifications. Customers want assurance that their providers are adopting cutting-edge technology and understand how to use it.

Second, our applied AI team works with partners to infuse Claude into existing solutions or build AI-native ones. Partners are upskilling employees, supported by our ecosystem and direct training.

Third, we work together on go-to-market initiatives.

Infosys has infused Claude into its Topaz solution. TCS has developed a solution through which it is taking TCS iON to regulated segments. Cognizant has skilled more than 30,000 employees to serve regulated sectors such as healthcare and insurance. Our priorities are adoption as a customer, building replicable solutions for joint go-to-market efforts, and ensuring production readiness and skilling. Our larger ambition is to take solutions created with GSIs in India to global markets.

 

Q: How much of Anthropic’s India strategy focuses on enterprise sales versus enabling developers to build products on Claude?

These are two sides of the same coin. Enterprises run software, while developers build it. Developers work across start-ups, digital natives, GSIs, IT services and industries.

Swiggy and Instamart, for example, created a solution that serves queries across 40,000 products. NPCI created AI Next, built as an MCP that can connect with other merchants. Razorpay created Merchant Studio, an agent through which users can connect not just to agents but also reclaim cards and undertake forecasting.

Enterprises are also running and developing these products. Manufacturers such as Reliance, Mahindra and Godrej are improving internal productivity and efficiency while building products for customers.

Banks, including Kotak, IndusInd Bank and Axis Bank, are creating intelligence layers and AI-led platforms. With inferencing available in India, governance and other parameters are built into the environment, enabling innovation for internal and customer-facing applications.

 

Q: How does it feel to lead a global company in India as a woman, and what strengths do women leaders bring to the AI frontier?

It is a moment of immense pride, and I feel grateful and honoured by the opportunity.

India has experienced four waves: the internet, mobile, cloud and AI. That experience matters when serving enterprises that need solutions that are safe, reliable, simple and integrated. The learning and credibility built over many years are important, and I believe this creates opportunities for women leaders who have lived through these eras.

Women leaders at the frontier bring several strengths. First is their range: solving mission-critical enterprise problems, working with governments on population-scale challenges, and collaborating with public-benefit organisations and NGOs addressing nuanced needs at the grassroots.

Second, women are good at bringing diverse teams together, adding different textures, colours and richness to how AI is formed. Third, they can connect the dots, close the loop and see the bigger picture.

In everyday life, women wear multiple hats. One day, you might be the captain of a ship, steering it through complexity and staying on course. Another day, you might sculpt things, create opportunities and make things happen. A third day, you might roll up your sleeves and do the grunt work because it needs to be done.

Women often do all three on the same day. I believe steering the ship, sculpting things and being a doer are ways women can help bring things to the frontier.

 

@palakagarwal64