Why ServiceNow isn't worried about 'Claudification' and sees AI as an opportunity

ServiceNow President and COO Amit Zavery believes AI will create a divide within the software industry, rather than disrupt every company equally. He argues that products deeply embedded in enterprise operations will be harder to replace, even as AI models become better at writing software. Zavery also explains why ServiceNow sees India as an increasingly important market and innovation hub.

By Shereen Bhan

Artificial intelligence has raised a difficult question for the software industry: if AI can write software and perform tasks that once required specialised applications, what happens to the companies that built those applications?

ServiceNow President and Chief Operating Officer Amit Zavery believes the answer depends on what a software company actually does for its customers.

In an exclusive interview with CNBC-TV18, Zavery argued that AI is increasing demand for ServiceNow products because the company's software is deeply integrated into enterprise operations. He said investors are too often treating software companies as one group without considering how AI affects each business differently.

That is also why he is not particularly concerned about the so-called "SaaSpocalypse" or the prospect of ServiceNow being "Claudified" by AI companies such as Anthropic and OpenAI.

Why ServiceNow sees AI as an opportunity

Zavery's argument starts with a simple distinction: companies do not buy AI merely because it is AI. They want technology that delivers a business outcome.

"You're not buying AI for the sake of buying AI," he said. The relevant questions, according to Zavery, are whether a product makes a customer more efficient, saves money or helps increase revenue.

If it does, customers will continue to want it. But a software product that performs the same function without delivering a better outcome becomes easier to replace.

Zavery said ServiceNow is already seeing stronger demand for its AI products. The company had initially planned for $1 billion in AI annual contract value, or ACV, but is now expecting $1.5 billion.

He believes that is one reason investors need to look at software companies individually rather than assuming AI will have the same impact across the sector.

Why some software companies are more exposed

Zavery does not see AI as the only factor determining which software companies will survive. He believes the bigger risk is failing to adapt when technology changes.

He pointed to previous shifts such as cloud, mobile and the web. Companies that did not change how they operated and used those technologies eventually suffered, he said.

He also sees a particular vulnerability in products built around a single feature that can be easily replicated.

"If you're integral to an enterprise, or integral to an organisation, or integral to a company, it is very hard to replace," Zavery said.

That is where he believes ServiceNow has an advantage. Its software, he argues, is not simply providing an isolated feature. It is connected to enterprise workflows and the way organisations operate.

The question, therefore, is not simply whether AI can reproduce something a software product does. It is whether it can replicate the wider system around that product.

Why Claude does not necessarily threaten ServiceNow

That distinction becomes important when looking at the rise of AI models such as Claude.

The concern around "Claudification" is essentially whether increasingly capable AI models could allow companies to build software themselves and reduce their dependence on established software platforms.

Zavery's answer is that AI models provide intelligence, but not necessarily the enterprise context and execution capabilities needed to run a business.

"They're providing intelligence. They're not providing the execution engine. They don't have the context," he said.

That is also how Zavery views ServiceNow's partnerships with companies such as Anthropic and OpenAI. Rather than seeing these companies only as competitors, he believes there is an opportunity to work together because each provides a different part of the technology stack.

Zavery also argued that large language models account for only a small part of ServiceNow's own intellectual property.

"If you look at the IP inside my product today, it's less than 10% LLM," he said.

The rest, according to Zavery, has been built over years around workflows, business requirements and the context behind enterprise decisions.

That is the part he believes cannot simply be replicated by giving an AI model the ability to write software.

Writing code is not the same as running enterprise software

Zavery's other key argument is that software development is only one part of the challenge facing an enterprise.

He estimates that writing software accounts for about 20% of the cost, while the remaining 80% involves maintenance, governance, security, compliance and upgrades.

That distinction is central to his view of the threat from AI coding capabilities.

"Claude can write you the software, but who's going to do the rest of the maintenance?" he asked.

In his view, an AI model producing software does not automatically eliminate the need for the systems and processes required to run that software inside a large organisation.

That is why he believes deeply embedded enterprise software is less vulnerable than products whose functionality can be easily replicated.

Where India fits into the strategy

India is an important part of ServiceNow's expansion and AI development.

Zavery said India could become one of the company's top three global markets and is already its fastest-growing market in Asia Pacific. It also has ServiceNow's second-largest employee base outside the US, he said.

The company has invested heavily in product engineering in India, while also expanding its go-to-market operations and partnerships.

Some of the AI work is being done in India. Zavery said more than 1,000 customers are using AI Control Tower, a product introduced last year, with a significant amount of its capabilities developed by the Indian team.

The product is designed to help companies discover and manage their AI investments, agents and systems, while addressing costs, governance and security.

ServiceNow is also developing capabilities such as Autonomous Engineer, AI Workflow Factory and voice capabilities for the Indian market, including different languages and accents.

Zavery said some of the products developed in response to Indian requirements can subsequently be taken to other markets.

Businessnext is one example. ServiceNow has invested $40 million in the business, which has primarily catered to India's financial sector, and Zavery said the capability is now being taken to Southeast Asia and Australia.

So what does AI mean for ServiceNow?

Zavery's view is not that AI will leave the software industry untouched. His argument is that the impact will differ sharply between companies.

Software businesses that fail to innovate, or products built around easily replicable features, could come under pressure.

For companies whose products are deeply integrated into enterprise operations, however, AI can become a way to deliver more value to customers.

That is the distinction at the heart of ServiceNow's argument.

AI may make it easier and cheaper to write software. But, according to Zavery, that is only one part of what an enterprise needs. The larger challenge is putting that software into a business environment and managing the workflows, security, governance, compliance and maintenance that come with it.

That is why he does not see the rise of Claude and other AI models as an existential threat to ServiceNow. He sees them as another technology that can be incorporated into the enterprise software stack — and potentially a driver of greater demand for what ServiceNow already provides.

This is the edited excerpt of the interview.

Q: Let me start by asking you about why the decision to bring your first-ever World Forum to India.

Amit Zavery: I think India is one of the global big markets for us. I think India is going to be the top three market for ServiceNow globally in general. We're seeing a huge amount of traction and demand around ServiceNow products, but in AI in general, the work which is happening in India is so interesting and innovative, and we have a large presence.

Q: It is your second-largest employee base outside of the U.S.?

Amit Zavery: Yes, for sure. Yes, and it's growing very quickly as well. The big thing we wanted to do was to bring our top customers together at one place, talk about the innovation, talk about the way we can help them, how we can partner with them, including a lot of the partners who are here who work with global customers as well as the local companies. So this was a great time to do it, and the demand as well as the response has been great. So very excited to be here and be able to talk about it.

Q: So you said India top three by when?

Amit Zavery: Hopefully soon. You know, India is growing very well. For us, we are very happy with the success we're having here. We have invested aggressively in product engineering for sure for many years. Over the last many years, we've grown our go-to-market capabilities, our partnerships, and the results are exceeding our plans. So we feel very good about the future, and that's why we continue to do this more and more, and continue to bring a lot of innovation out of India as well as for India.

Q: Okay, I do want to get a little bit more specific with you, though, on what the India contribution is. You don't disclose that country-wise, but how much of your revenue comes from India is driven by India today.

Amit Zavery: Well, I think if you look at a lot of the business we're getting in India, it is by Indian customers. So we work with all the large banks here, a lot of the large telcos, public sector, both state, local, as well as the federal government, the central government, so that's been a very big part of our business. Healthcare is growing very fast as well, so we don't have specific numbers in terms of how much it contributes to our total thing, but the growth has been a big part of India. So if you look at Asia Pacific specifically, the fastest-growing market for us is in India, and we have— that's why we increased the number of people who are based out of India working with the local teams. Even if you look at the partners we have, the system integrators, others, they're doing a lot of work inside India itself, not just outside for multinational companies.

Q: I'll come to the system integrators in just a second. But you've had some announcements that you've made, which are specific to India. Let's talk a little bit about the plans as you move forward.

Amit Zavery: I think some of the things we've done - we have this concept of AI Control Tower. We have a product we introduced last year, 1,000-plus customers. A lot of the work was done in India here. The Indian team has been building a lot of the capabilities, and the key idea is that how do you go and discover all your AI investments, AI agents, AI systems you're running, make sure that none of the systems are doing rogue things or going haywire and destroying your company, as well as making sure that you're cost-effective, governance, security, all that stuff. They have become a very central tenet to anybody who wants to adopt AI. So that's been a big part of it.

And then around that, we've been building a lot of other capabilities. For example, this idea of autonomous engineer. How do you build an application very quickly using vibe coding, but also with all the governance and security built into a platform as well? The idea of AI Workflow Factory, where you have a whole life cycle of what it takes to run a workflow, how do you manage it, how do you operate it, and then also connect onto various different systems and get the return as required.

So we continue to innovate out of India, and a lot of these things, like voice capability building for the Indian market, so that the people can interact with it with different languages, different accents, and other things like that. So, tonnes of things going on. We can go on and on in terms of the innovation here, but it's been exciting times.

Q: You know, it has been exciting times. You've done a $40 million bet as far as Businessnext is concerned, and that largely has catered to the financial sector in India, banks particularly. Anything more on the anvil?

Amit Zavery: No, I think we have our hands full, and we're doing a lot of things, as you can see. But this idea of Businessnext is a very good example, right? We're doing things that are much more specific to a particular market because we're seeing a specific demand, and that can be replicated outside India as well.

So, Businessnext is a good example where we're taking that capability to Southeast Asia, to Australia, and they have a lot of expectation about what we can do jointly. We have a lot of expectation in terms of what they bring in some of the domain, and a lot of the customers we work with jointly are getting a lot of value. We identifying partners like that, who we can really co-innovate, as well as quickly take to market and build the business for them with us?

Q: You know, let's now talk about what's happening as far as the world is concerned, and of course, this AI-driven disruption has captured mindspace headlines, and certainly, you know, it has put pressure on stocks like ServiceNow. 2026 - how would you describe the year? I mean, you know, while quarter after quarter on the financial front, you're actually meeting your estimates, or in fact beating them, you've raised your revenue guidance as well, and yet the market is punishing your stock. How do you explain that?

Amit Zavery: I think it's very hard to understand investor sentiments, depending on not them understanding nuances in the business.

Q: What are the markets getting wrong about the story?

Amit Zavery: Quite a few things. One, I don't think they understand how AI is going to help a lot of companies like ours. They're kind of putting every software company in the same bucket. We are really getting a lot of acceleration in our revenue and our customer demand because of AI.

If you look at our revenue numbers, we talked about our AI products. Our plan was $1 billion ACV. We are going to be $1.5 billion. We raised the guidance. That's 50% on top of a large number itself, and we continue to see the acceleration.

So I think the understanding of how AI helps specific companies versus others, I think it has to be a little more nuanced, because we are really the operating system which is connecting various different things inside an enterprise. To try to replace or change that, it becomes very difficult because we are really adding value, making companies operate better, and be able to get results out of it.

So, products are really driving value and results. That's why demand is going up. Our results are great. We're helping really innovate with AI with our customers on the same platform. So, those things are all working out.

So, I think the confusion for many investors is not understanding what it does to whom, and that's why they're putting everybody in the same [bucket]. But if you look at the last few months, we seem to be getting a message through.

Q: There has been some recovery from the lows that you've seen, but still nowhere close to your 52-week highs. But you know what I want to understand is, you said the market is tending to put everybody in the same basket. Now, if you believe that you're not vulnerable, who do you think is vulnerable?

Amit Zavery: The ones who are not going to innovate. When any transformation happens in technology, you saw that with cloud, you saw that with mobile, you saw that with web. If you don't innovate, doesn't matter what technology it is, you're going to suffer.

So I think it's companies who are not going to change how they operate and how they use this technology to change, make a difference in the product. They're going to have a difficult time.

Second are the companies who are going to be very feature-oriented. They have a vertical stack providing one particular feature that can be very easily replicable, because they can be replaced by buying something else somewhere else because you're not integral to the enterprise. If you're integral to an enterprise, or integral to an organisation, or integral to a company, very hard to replace.

Q: But how is embedded software in the AI era then very different? Because we've been doing exactly that, isn't it?

Amit Zavery: No, I think, see, AI software has to be in the product to really drive the difference. But you're not buying AI for the sake of buying AI. So, am I adding value? Am I making you more efficient? Am I saving you money? Am I helping you grow top line?

If you're going to answer the question in the positive with your product, customers want you. If you're going to be the same product doing the same thing without any difference from the outcome perspective, that is very interchangeable.

Q: The two nightmarish scenarios that were painted at the start of 2026: one, SaaSpocalypse—is it all over as far as SaaS companies are concerned? And two, will you be "Claudified"? I want you to explain your thoughts on both, because you have multi-year partnerships with Anthropic, with OpenAI, and again the question is, are you actually working with your competitor in the garb of your partners at this point in time? Will they actually be an existential threat to you a few months, perhaps down the line?

Amit Zavery: If you look at what they're doing, right? They need help from enterprise companies who know how the domain works. They're providing intelligence.

They're not providing the execution engine. They don't have the context. So why do you think they're coming to me every day and say let's partner together and help us get to market. But they are a very small part of the full stack.

If you look at the IP inside my product today, it's less than 10% LLM. The rest is all IP I have built for years, understanding how workflows work, what are the business requirements, how are the contexts associated, why the decision was made. Those things are not replicable by any LLM.

So I think it's a very good opportunity for us to partner in some cases. But we're also very smart about what we partner on. I mean, of course, everybody has ambition. So obviously, what has happened with AI - the lines between different teams have blurred. Everybody can claim they can do everything. I can write the software, the problem is you write software without domain, it's useless.

The software part of writing, cost-wise is 20%. For an enterprise to maintain the software is the other 80%. Understanding the governance, the security, the compliance requirements, maintenance, upgrades is 80% of the cost. That is not easily replicable just because you have Claude. Claude can write you the software, but who's going to do the rest of the maintenance?

So that's why it's very difficult to replace things we do, and that's why we are not really worried about it. I mean, customers are not coming to me any day. I talk to CIOs. Nobody says I'm going to go and rebuild this thing, ever.