Synopsis
Indian organisations, from large public sector players to SMEs, are engaging more actively with artificial intelligence. They are moving beyond proof-of-concept stages to scaled deployments, said Irina Ghose, managing director, India at Anthropic, on Thursday.Listen to this article in summarized format
She said companies are increasingly deploying AI for real-world applications and prioritising trust and reliability, not just model capability.
“When you are building, think about solving real problems in India, especially last-mile challenges, while also contributing to broader global outcomes. The opportunity set is immense,” Ghose told developers while speaking at vibe-coding AI startup Emergent’s hackathon event.
The startup is backed by Khosla Ventures, SoftBank and YC, and is valued at $300 million after a $70 million round in January.
Anthropic, which is now attracting investor interest at an $800 billion valuation, said it counts India as a key market and is seeing faster adoption cycles, with enterprises no longer questioning the relevance of AI. “They feel the bus is arriving and are deciding when to get on,” the MD said during the fireside chat, adding that organisations no longer need to justify why AI, as that question is already answered.
Ghose said India’s strengths lie in its diversity, scale and ability to solve last-mile challenges. “If you can solve these constraints, you can build solutions that are relevant not just for India, but for many other parts of the world.”
She urged builders to focus on harder problems rather than entry barriers. “Focus on solving the biggest, most complex problems, the ones that haven’t been solved yet.” The India MD told developers that the entry barrier has already come down significantly.
Selective approach to Mythos
On initiatives such as Project Glasswing, announced by Anthropic in early April as part of its latest model Claude Mythos Preview, the company chose not to release the model to the general public.
The company formed a restricted consortium of around 40–50 organisations, which have been given early access to the model to test and secure their systems before similar capabilities become widely available.
“The goal is to test and validate capabilities in a controlled way with a set of partners that provide broad coverage,” Ghose said, adding that the idea is to ensure that the performance and capabilities of models, especially in areas like cybersecurity, are robust.
“As we learn from the initial set of partners, we will expand the scope in future phases.”
She added that Anthropic has already released supporting resources for developers and will expand the scope in future phases. “We have already published capabilities around AI for cybersecurity through blogs and white papers. These are available for developers to explore and implement even now.”
The company is aligning with sectors prioritised by governments, including education, skilling, agriculture and healthcare. Ghose said Anthropic is building “from India, for India,” in collaboration with developers and the broader ecosystem.
On timelines for artificial general intelligence (AGI), she said timelines are compressing, given how fast models are improving. At the same time, as capabilities increase, it is equally important to ensure responsible deployment.
