AI is already transforming government, Maharashtra Chief Minister Devendra Fadnavis said at the World Economic Forum’s panel discussion titled ‘AI for Vikshit Bharat’ moderated by Srutijith KK. “We (Government of Maharashtra) are now trying to incorporate AI into the entire process, governance and service delivery.”
Fadnavis said India’s digital public infrastructure is an equalizer. “Now is the time to leverage AI to leverage this digital infrastructure for the greater public good,” he said, referring to the state’s agriculture initiative, adding, “We created AgriStack…the entire data of land records, crop records, everything, digitized for every farmer.”
Fadnavis also highlighted Maharashtra’s plans to build a 200-acre innovation city and position it as a hub to attract AI-driven investments, startups and talent.
Sanjiv Bajaj, Chairman and Managing Director, Bajaj Finserv, called AI a disruptive but familiar technological change. “Whenever there are discontinuous technological innovations, and AI is clearly one of them, there is a big change. At first there is hype and fear,” he said, drawing parallels with the earlier transition from steam to electricity and the rise of the internet. “Over the last 200 years, the world has become more productive and richer,” he said. He outlined three stages of AI implementation: productivity, effectiveness, and innovation, adding that some group companies’ AI-driven call centers have already achieved “30% productivity gains.”
At Bajaj Finserv, AI is already reshaping advertising, customer engagement, and lending. “We produce thousands of marketing videos every year, all created end-to-end by AI,” Bajaj said, citing a Diwali campaign as an example where “we customized around 300,000 individual ads across our stores in 15 days.” As for loans, he said AI bots are currently negotiating loans in Hindi, English and mixed languages. “Out of 4.5 million loans per month, we have already executed 30,000 to 40,000 loans end-to-end with our AI products,” he said.
While AI adoption is clearly accelerating, Sanjeev Krishan, chairman of PwC India, said there is a significant gap between adoption and value creation, citing findings from a global report released at Davos. “Only 12% of CEOs say they have benefited in both sales and profits from using AI, and that’s largely because no one sees AI as a tool to revolutionize their companies,” he said. Krishan argued that while AI disruption is inevitable, the real challenge lies in preparing for it. “Humanity will always outpace any technological development, because at the end of the day we are the ones innovating.” The issue is not about job losses, but whether people have the right skills, he said.
Krishan said India needs to urgently rethink its education system if it wants to remain relevant in an AI-driven economy. “The higher education system needs to be restructured if it wants to be relevant to what happens next.”
PwC launched the ‘AI Edge for Viksit Bharat’ report at ET House in Davos.
Zerodha co-founder and investor Nikhil Kamath cautioned against applying traditional valuation metrics to AI companies.
He said that with the exception of companies with established businesses such as Alphabet and Nvidia, “most AI companies don’t have the revenue to warrant such multiples.” “You can’t value companies in the AI space today based on the revenue they earn today,” Kamath added, arguing that current valuations are extrapolations of an uncertain future. “No one knows, so anyone can bet,” he said.
Kamath said India should avoid replicating Western AI models and instead focus on building applications that surpass core models. “The mistake I don’t think we should make is trying to copy what Western companies are doing because they have more risk capital on hand,” he said.
He also warned about platform dependence. “Now is the time to stop relying on one platform,” Kamath said, appealing for diversification. Looking ahead, he added that “contrarian behavior and nuance… will become increasingly important,” noting that originality is difficult for models to replicate.
