Artificial intelligence adoption is accelerating rapidly in India, with enterprises moving beyond experimentation to large-scale deployment, according to Irina Ghose, Managing Director for India at Anthropic.

“India is not just a large market for us, it is shaping how we build and evolve our products... India is now our second-largest user base globally, and what stands out is the depth of usage,” Ghose said at the Fortune India Startup Summit 2026 in Bengaluru. “Nearly half of users here are using Claude for technical and development work, which is significantly higher than many other markets.”

She added that India accounts for around 6% of global conversations on the platform, reflecting strong engagement, while about 12% of usage is driven by research and institutional applications.

To tap into this momentum, Anthropic has expanded access to its latest models across 10 Indian languages. “Localisation is critical for us, not just for access, but for relevance in real-world use cases,” Ghose noted.

Outlining the company’s India strategy, she said it is focussed on three areas: enterprise adoption, the builder ecosystem, and last-mile access. “We are working across industries like BFSI, manufacturing, digital-native startups, and even the public sector, while also enabling builders—from IT services firms to startups—to scale AI adoption,” she said.

Early trends indicate a clear shift in enterprise behaviour. “We are moving beyond basic productivity use-cases. Enterprises are now looking at complex, mission-critical applications,” Ghose said.

Large IT services firms are leading this transition. Companies such as Cognizant have deployed AI tools at scale and integrated them into internal platforms to improve delivery efficiency. In aviation, Air India is leveraging AI to enhance customer service and response accuracy, while platforms like Swiggy are building more seamless user experiences using AI.

Highlighting the next phase of adoption, Ghose said, “Agentic AI is the new frontier. We are seeing enterprises move from conversational AI to systems that can actually execute tasks.”

However, she acknowledged that challenges remain in scaling adoption. “The three things enterprises consistently ask for are trust, reliability, and change management,” she said. “They want systems that are accurate, that don’t hallucinate in critical domains like finance or legal, and that integrate seamlessly into their existing workflows.”

She emphasised that performance alone is not the deciding factor. “The best-performing model is not necessarily the winner. What matters is whether enterprises trust the product and the philosophy behind how it is built,” Ghose said.

On adoption trends, she noted that while startups and digital-native companies continue to move faster, traditional industries are catching up. “Today, it’s the builders within organisations who are driving this shift, across banking, insurance, manufacturing, and more,” she said.

Fintech and IT services, in particular, are emerging as strong adopters. “In sectors like fintech, AI is not just about efficiency; it’s about creating differentiation,” she added.

Addressing concerns around job disruption, especially in IT services, Ghose reframed the narrative. “The fear is understandable, but the framing is often wrong,” she said. “AI is making coding and development far more efficient, but that efficiency is being reinvested into solving more complex problems.”

On safety, she reiterated Anthropic’s approach of balancing innovation with responsibility. “Capability and safety are two sides of the same coin. You cannot build one without the other,” she said, adding that the industry is still evolving in defining fully trustworthy AI systems.

Founded by former OpenAI employees, Anthropic is a San Francisco-based public benefit company backed by investors, including Amazon. The company recently expanded its India presence by opening its second Asian office in Bengaluru and appointing Ghose, formerly with Microsoft India, to lead its operations.