Global Uncertainty Tests India’s Growth Story

Live from the New Economy Forum in Delhi, this is Insight with Haslinda Amin. One of the world's fastest-growing major economies competing globally for investment, technology and talent. As geopolitical tensions reshape trade and AI transforms industries, can it lead to sustained long-term growth? We will focus on all that.

You're at the Bloomberg New Economy Forum as we count down to the keynote speech from Narendra Modi. Investors and business leaders will be listening for clues on India's economic ambitions in the changing global order and plans to attract investment at a time of heightened uncertainty. Also ahead, we will speak live with the IMF chief economist as the fund raises next year's global growth forecast but warns resilience is under strain. And Mike Cannon-Brookes joins us exclusively on what the transformation means for customers, competition and growth. We will discuss India's growing role in durable healthcare and the opportunities ahead. All that and more coming up as we bring you the biggest names and best conversations from the Bloomberg New Economy Forum.

Bloomberg's New Economy Forum is just getting underway here in New Delhi. India is hosting for the first time. Later, Prime Minister Narendra Modi is set to address key policymakers and business leaders gathered here. Let's cross over to my colleague Shery Ahn. This is the eighth edition of the NEF, hosted for the first time in India. What are we expecting, Shery?

Prime Minister Modi is expected to arrive anytime now at the hotel here in central Delhi. We've seen participants of the New Economy Forum already lining up to go see the Prime Minister. He already welcomed Bloomberg, saying India is not only an investment destination anymore, it's actually a reliable partner in shaping the global future. We are seeing heavy security around the hotel right now, at the entrance of the hotel, but it's not just because of Prime Minister Modi's arrival, it's because of all those high-profile guests that you mentioned will be arriving here for the event. The former Commerce Secretary of the United States, not to mention the Finance Minister of India, also corporate leaders as well, and global leaders from Brazil, all coming here to India to have different conversations around the state of the Indian economy and the global economy, not to mention leaders of geopolitical conflict we are seeing around the world.

It's not just the security because we're expecting all those newsmakers to come. We've seen security, police and paramilitary forces around Delhi in the past few days, and tensions have been rising. We're talking about student protesters being detained and released over the weekend. We have the Janata Dal party vowing to continue its protest, seeking electoral reforms, and they want the resignation of the chief electoral commission official on allegations of manipulation of voters. There is a lot to contend with in this backdrop as we're waiting for all these newsmakers together here in New Delhi.

As you said, the Prime Minister is expected to speak soon against the challenging backdrop. It's not just about the protest, but also the economy that has slowed down from 7.7% to 6.5%. What is he expected to say, Shery?

Prime Minister Modi has already welcomed the New Economy Forum, saying he wants to build India into a reliable source of growth for the world economy. The challenges are there, relative to the entire world. At the same time, inflation is accelerating. Inflation last month came in with an acceleration of 5.5% from a year ago. This is toward the higher part of the tolerance band from the RBI. Perhaps not surprising given the issue of energy. 90% of its oil needs come from overseas, not to mention the expectations of higher rate hikes. Inflation expectations going up. This morning it was rainy and wet, but that hasn't necessarily been the case for this monsoon season, which delivered about 13% less rain in this country. That's feeding through to crops and food prices and monetary policy. The RBI started its hiking cycle recently given these concerns.

That's right, and the rupee is hovering close to a record low, 97 to the dollar. Shery Ahn at the New Economy Forum here in Delhi. The IMF says global growth should pick up next year with a boost from AI investment. This morning, the world economy is nearing the limits of its resilience in its latest World Economic Outlook. The IMF maintains a 3.4% growth forecast for 2027 and has raised its expansion outlook for this year to 3.1%. Businesses have continued to adjust to repeated shocks, but warns it could change as the energy squeeze persists and financial risks continue to build. Let's get more on the IMF report with Avril Hong, who is standing by in Bangkok. Avril.

A very special guest indeed, the chief economist of the International Monetary Fund. Great to have you with us. Talk to us about what you're seeing in the global economy, because there is resilience. The question is, can it continue into next year and beyond?

Thank you. Indeed, the global economy has been resilient, but that resilience is reaching its limits, with risks tilted to the downside. We expect the global economy to grow this year, 40 basis points lower than last year, and to pick up again to 3.4% in 2027. There were two main factors. First is AI, which has led to demand, pushing activity and prices up. The other element is flexibility in energy markets, which has meant adjustment, rerouting of supplies, increased production outside the conflict region, and substitution toward other sources of energy and renewables. All these margins have had to be resilient, and some of them are reaching limits.

Exactly the point that has been raised in the report, in the way that energy inventories have been drawn down. And with that in mind, do you think AI will continue offsetting the shocks that could come to growth? What is in the balance here?

There is a lot of uncertainty on AI. We expect the AI cycle to continue for some time, and that is pushing investment. The key question is whether investment will crystallize into higher productivity gains. We haven't seen those productivity gains yet in the macro data. There are some studies that are more granular, looking at micro gains for productivity, but we haven't seen those reflected yet.

What about the impact from AI on price pressures, beyond productivity, but also on how prices are fed through into an inflationary aspect?

So far, we are seeing only the demand effects of AI, the increased investment. That is pushing up prices of key products, including chips and semiconductors. So that is the source of inflationary pressure in many countries.

Talk to us about what you're seeing in terms of the indicators that AI is working to improve productivity. What will the IMF be watching out for?

Those are not yet showing anything from AI or any significant change in productivity. It might still come, it is early on. What we are seeing right now is increasing investment, increasing trade, on how AI is driving this demand.

What about when it comes to the era of higher yields? What is driving this, and how much of it is structural?

There are three elements behind the increase in yields. First is AI, activity and prices, and the response from central banks. The second is energy prices, commodity prices triggered by the war, increasing commodity prices and the global supply shock. It drives up prices and hampers productivity. The last element is an interruption on premium terms, with higher interest rates. Investors require an extra premium or compensation for the risk. That is playing out as well.

What is it all going to mean for emerging markets? Spreads are tight. At some point, do you get the sense that things might turn?

Spreads have not moved in emerging markets, but obviously the base rates have moved significantly. So that is going to be a significant downward hit to demand in emerging markets.

Thank you so much, great to chat with you. Thank you again for your time today. The chief economist of the IMF.

Good to see you, Avril Hong on the sidelines at the IMF and World Bank meetings in Bangkok. Still to come, Mike Cannon-Brookes joins us exclusively on the company's push to AI, the challenge from ChatGPT and Claude, and what the transformation means for enterprise software. Keep it here with us. This is Bloomberg.

Welcome back. We're coming to you live from the Bloomberg New Economy Forum here in New Delhi. A sharp selloff in global and chip stocks is underway. The Kospi plunging 7% as we speak, and investors reassessing the outlook for the AI boom. The latest trigger: a warning from a CEO that financing, not chips, may become the biggest constraint on AI expansion. Let's bring in our markets reporter, Anthony Stephens. Put it in perspective for us, Anthony.

It's a tough day across the Asian semiconductor space. It follows the weak session in the U.S., into hyperscalers, banks and consumers. Asia is heavily geared to the big capex base, most notably in Korea. To a lesser extent, Taiwan contributing very much, getting a sell-off, down 2%, just shy of the 8,000 number locals were waiting for. Agonizingly short of the number, just as local leverage hit records. They will absorb foreign selling. Look at the board, and you can see another market in a different place, and that is India.

India has been kind of the negative. We will not be seeing this kind of drop in India. The overnight indication is flat, despite a high inflation number at 5.5%. We are well above the RBI's, but we did outperform in the ADR session overnight. That's because India has been lagging for so long as the economy looks to find its footing in the new AI paradigm. You can see how badly it has done. This kind of laggard play will be watched in the earnings season. We see it close a little bit, and that could be coming into play in the back half of this year.

Our markets reporter Anthony Stephens there for us. Half a world away, Australia's technology sector is grappling with many of the same issues as Anthony was talking about. Let's take a look at how they are trying to get a better advantage. A co-founder and CEO spent more than two decades shaping how AI works. He joins us now exclusively from the Lion City. Good to have you with us. Let's talk about that. When it happened, you used $250 million of your own money to buy back your shares. What do you think Wall Street misunderstood about the whole situation, and how vulnerable companies really are?

Thank you very much for having me. We are doing extremely well. We have accelerated for a number of quarters in a row. Our AI-driven platform is really being taken up by customers all around the world. I'm here in Singapore visiting a number of customers across the region, and our solutions, our products and platform help companies to adopt AI in terms of how it comes out into their teams, into their knowledge workers, and actually becomes part of their business processes in the ways they work. Customers that use our AI solutions grow at twice the rate of those that don't. So I think we are justifying that AI is a huge tailwind, and one of the best things ever to happen to our company is the new technologies and tools we have to solve the same problems for our customers' teams in better ways.

The thing is, Claude, an agent like Claude can build custom workflows. Why should companies be paying you millions and millions of dollars?

There's a lot of agentic harnesses that are very good. I use one of them myself to help us build and improve our software applications. Fundamentally, it helps teams of people collaborate in organizations all around the world. Service teams, business teams, strategy leadership teams and technology teams, those teams of people use many different applications, put them together, and what we have is a big part of our platform, which is the context graph of enterprise knowledge on how we take all the history of decisions and documents and information of the company and put it together into a context graph so we can get better answers to Claude or any agentic harnesses being used.

So we have a big partnership with Anthropic, OpenAI, and Google, bringing the enterprise knowledge of those processes and the history of decision-making. If I can't understand why you made a decision as a team or group in your company in 2019, you can now make the best decisions in 2026 or 2027. Your agents certainly can understand that. So we help companies take that information and connect it into the teamwork graph and provide a tool with agentic harnesses, while at the same time having a human experience of how agents and people will work together.

Mike, you're navigating one of the most important times for the company. You recently cut about 1,600 of your workforce, about 10%. What was the tipping point that led you to make that decision?

Obviously, those are always very hard, challenging decisions. Part of the job is navigating and making the right decisions. In that case, we needed to rapidly change as a company. It was very clear at the time. The AI part of our business is doing very well and growing very fast. The enterprise part of our business, dealing with the world's largest companies like Bloomberg, is a fantastic customer. Many other very large customers around the world are growing very fast. So we needed to change inside the company to invest even more heavily in those areas that are doing really well in terms of AI and the skill base. That is a part we are still growing as a company incredibly strongly.

We grew 29% revenue last quarter on an annualized revenue rate. Even at the scale we are at, we are accelerating as a company and continue to do so due to the AI tailwinds we have in terms of how we're helping companies and their teams be more effective across all the business processes.

I want to take a look at the tech space as a whole, the IPO market in particular. What is your reading on it? You've been quite vocal on the poor quality of tech IPOs.

I don't know all the details. I tend to stick to my lane and things I focus on, and obviously we are here dealing with customers as a good example. I think companies that have a really deep technological foundation are able to navigate these times and these changes. Certainly, the less we spend time focusing on our technological foundations over the last three or four years now in the AI era, the approach we're taking is we have to create our way through it. We have to create new technologies and leverage all these technologies to build new value for customers with all of these amazing tools that we have access to in terms of AI and the models and how we bring it to customer value. I would say it's about having a deep technology core, at least for us. Other companies can do their own thing.

It's pretty interesting that the fund managers at home balked at the valuation of $44 billion. Is this a healthy development? People are beginning to question the AI space and the tech space and the amount of money that's been plowed into that space.

Again, I can only comment from the viewpoint of Atlassian. It's fabulous to see their support throughout the year as we've gone through interesting market times as a business ourselves. At the same time, it's wonderful to see the understanding of the products we've built among both the Australian fund management community and across the U.S. investors. I spent more time with European investors. I think the attitude to technologically driven software companies in the AI era has changed because the economics of software is still a fantastic business. All of that lets us grow customer value faster, and we're accelerating our business for that reason, because of our investment in technology and delivering for our customers. It's great to see the investment community around the world valuing that.

That takes us quite nicely into what is hot and what is not. There have been calls for us to slow down AI development. What is your take on that?

From our point of view, we love model progress. Our job is to take more progress and harness it for customers. We are building in the software layer, we are building solutions for customers that take the AI model progress and harness it for the value of the customer, solving the problems that the customers have. It's our job to combine different models. We have more than 75 different models in production to give the customer their own set of choices and different values. Progress is a great thing because we can solve ever more problems, and as the models move forward, it's not just the latest models, the more mature models are a great thing for us at the same time because as a technology world shows up to cost-optimize all the models, we can deliver ever-increasing value at prices for our customers, which is really important. AI capital management is a concern for all customers. We let you reduce your token usage proportionally as it increases. So it's a very significant savings for our customers, accessing knowledge from any of those AI houses that are out there.

Mike, talk to us about your expansion plans. We know you have quite a big operation here in India.

We do, yes. Just past 40 staff here in Singapore and hiring strongly across the region. Southeast Asia and into continental and North Asia, all are big and very fast-growing markets for us. Singapore at the moment is one of our faster-growing markets in the region. India is one of our three big offices. We have staff in probably 18 or 19 countries around the world at the moment, and 26 out of the 27 states in India. We have a big office in Bangalore, but it has been a big success story of Australian-Indian cooperation, the two countries working closely together. We've had great success there, growing in the region, and we are there for the talent. It's an incredibly deep talent base. We do fundamental engineering. The things I just spoke about are built out of the Indian office in collaboration with Australia and America. We could really run a 24/7 operation around the world, finding the best talent wherever we can find it. India has a lot of fantastic technical talent, so it has been a real success story for us. I was there just last month.

I want to talk about the AI models. Increasingly, the U.S. is asking the world to choose between their models and the Chinese models. Perhaps enterprises will become more comfortable in using those models, the likes of DeepSeek. Pretty controversial statement to make at this point in time. What really are your thoughts on it?

I think we do believe in customer choice. So customers are seeing the value in AI beyond prototypes and pilot programs, deeply embedding it into their business processes. We see that every single day. As they do so, they have a new cost management exercise in their business workflow and technology teams, how they manage the AI costs that are rising. Some have very strong views on where models have come from. Others are more flexible and take advantage of blending different models. That's absolutely part of the concern. I think open-weight models across the world are picking up steam. We have a number of fine-tuned Nemotron models. I think open-weight models broadly are growing. We have a question around where those models come from and customer preferences of that. We see it shifting slowly as it becomes a cost equation and the content of the open-weight model is more understood. I think we will continue to be customer-led about that. Model progress, model competition is a great thing because we can deliver better solutions for our customers at a better price and more magical AI outcomes across the platform. We are excited by all of the model progress and competition happening to increase quality and bring prices down of every token.

But Mike, it is about compliance, national security. What do you tell governments and enterprise clients who are keeping away from the Chinese models because of these concerns?

I think it is totally their choice, if they understand the trade-offs they are making. They are making a trade-off around cost, risks, absolutely compliance, data sovereignty. Understanding where your data is being used as a company, meaning all of the regulations of your country or industry, always important and top of mind for customers. Our job with those platforms is to support all of those requirements, whether that is sovereignty and data resiliency in Europe. We just shipped inference residency last week at our conference in Amsterdam. The inference on all of those latest frontier models for European customers are being rolled out to be done in Europe, not being done in the U.S., even if the data was living in Europe. It is important for customer choice to be there. Over time, the model blending is becoming more and more important. You are seeing a lot of models taking an open-weight model, adjusting it, changing it, and rolling it out. We have to be careful about the stack of technologies we have, just as we do for any other areas, whether it is your operating system, is it open-source or not open-source. I think customers are making smart choices, being very thoughtful. Having had some conversations with the Singapore government here, the same thing. People are thinking ahead and trying to make thoughtful choices. It's exciting to see how it will all develop.