“We’ve been in business in India since 2001 and yet we didn’t have support for the majority of the language that the men and women in this country use. 53% of this country uses Hindi,” he said. “That changes now.”
For Afshar, the announcement is inseparable from a larger conviction. “You can’t lose by betting on India.” Stanford has rated India number one globally in AI readiness. The country produces three million STEM graduates annually, three times the United States figure. There are 161 unicorns. The digital economy is growing twice as fast as any other sector and is on track to reach a trillion dollars.
Salesforce has crossed one billion dollars in India revenue, employs over 14,000 people across six cities, and is expected to unveil a Salesforce Tower later in 2026. Its ecosystem is forecast by IDC to add two million jobs and $90 billion in new revenue to India by 2028, excluding Salesforce itself.
The Hindi announcement, he said, was not a product decision made in a boardroom. It was something agentic AI forced the company to confront. With agents absorbing customer support volume, Salesforce found itself with capacity and clarity it had not had before, and used both to look at where it had been underserving its markets. What the fifth of his five Rs framework, reclaiming latent value, surfaced was a language gap the business had been sitting on since the year it first opened its doors in India.
He also made a point about what the agentic moment means for India in structural terms. Historically, scale required resources. A large company could do what a small one could not simply because it had the headcount. Agents change that. “If you are a small business now you can compete with a large business because scale is not an issue. You have now digital labour.” For a market with India’s depth of entrepreneurial ambition, that is not a marginal development. It is a shift in who gets to compete.
Near the end of the India conversation, he said something that sat differently from the rest. “Some of the top leaders at Salesforce, frankly some of the top business leaders in the world, are from here.” It carried more weight in context, because earlier in the same interview Afshar had disclosed for the first time on air that he is an immigrant refugee, that his family experienced significant hardship, and that building everything from nothing shaped who he is. The line about Indian leaders was not a statistic. It was a recognition.
Afshar also confirmed that Salesforce has stopped building products on ethical grounds, more than once, without naming what. “If we don’t believe it’s for the betterment of society, we stopped building it.” Behind that decision sits an ethical and humane use office involved in every product feature, six dedicated research teams, and mandatory ethical and security training across all 85,000 employees. Agent behavior is tested across thousands of scenarios before it reaches a customer.
On whether AI agents can be trusted at all, given reports of agents resisting shutdown or manipulating users, Afshar said the answer is governance embedded in culture, not policy. “This technology is more powerful than any technology we’ve ever had in humanity. So if you’re not guided by a good culture, good core values, you could find difficulties.”
The AI dividend
South Korea’s government recently sparked a global debate, and a sharp market selloff, by floating the idea of an AI dividend: windfall profits from the AI boom returned directly to citizens. Asked whether he believed something like that was necessary, Afshar did not answer directly. What he did instead was more revealing.
He disclosed, for the first time on air, that he is an immigrant refugee. “My family and I experienced incredible hardship and it took, I would say my parents two decades to recoup as immigrant refugees who experienced that in their 40s. Everything changed and it was dramatic and very difficult.” He said the experience gave him zero sense of entitlement and made him permanently others-oriented.
It was a rare personal moment from someone who speaks almost entirely in frameworks. And it was followed not by a position on UBI or windfall taxes, but by Trailhead. Access, in his framing, is the dividend. Education is the redistribution mechanism. Three million Trailhead users in India. 150,000 university learners trained on agentic enterprise through partnerships with 3,000 colleges and universities across the country. Whether that answer is sufficient to the scale of the question being asked around the world is something he left open.
Three forecasts for the next decade
Afshar closed with three bets for the next decade, each one building on the last.
The first is digital twins of the world. Not just manufacturing facilities but airports, hospitals, banks, and retail outlets, entire physical environments mirrored in real time using platforms like NVIDIA’s Omniverse, so businesses can test and optimize before anything happens in the physical world.
The second is multi-agent orchestration. Today the conversation about agents is almost entirely about what they do for businesses. That framing, he argued, misses half the picture. Every one of the 5.5 billion people with internet access will eventually have their own agent. Consumer agents and business agents will interact, negotiate, and make decisions together.
“Figuring out where to vacation, where to send kids to school.” The infrastructure of daily life, running through agentic systems most people will never think about in those terms.
The third is ambient computing. Computing is getting smaller, closer to the body, and embedded in everything connected. Every connected object carrying reasoning, the ability to adapt, the ability to personalize. “Everything I can do with my Salesforce platform, I can do with smart glasses. I can just speak to it.”
“The combination of ambient computing, digital twin of the world, and multi-agentic orchestration are three big bets that will shape the next decade,” he said.
Technology, he said, is always last on the list. Culture first. People second. Processes third. “And finally technology.” For someone whose job is to evangelise one of the world’s largest technology companies, it is either a deeply held belief or the most sophisticated sales pitch in the room. Possibly both. “This is not about technology transformation. This is relational transformation. After all, the R in CRM is relationships,” he concluded.
