Enterprises must partner with consulting firms for AI: IBM's Mohamad Ali

IBM Consulting SVP Mohamad Ali says AI will be "net positive" as firms adopt human-plus-digital worker models and shift investments toward automation

IBM Consulting is crucial in the technology giant’s growth. According to IBM’s 2024 annual report, 80 per cent of its artificial intelligence (AI) bookings were fuelled by the consulting practice. The unit earned $20.69 billion as revenue in 2024, but growth has softened due to macroeconomic uncertainty. Mohamad Ali, senior vice-president at IBM Consulting, believes AI will be “net positive” despite the softness in demand. Ali, in a video interview with Shivani Shinde, talked about consulting and layoffs at IBM. Edited excerpts:

 

How has AI impacted consulting?

 

There’s a broader impact of AI on the global economy. The consulting space has been soft for some time, and I think that predates the big AI boom now. In the long run, AI will be a net positive for consulting, but there will be a lot of disruption in the middle.

 

If you look at history, every time there has been a technology inflection that lowered the cost of production, GDP has gone up significantly. I believe that will happen again. In technology consulting, a lot of what we do is write applications, develop software, and build systems. AI allows us to build systems and apps much more efficiently.

 

For instance, earlier the big focus was on massive applications, like SAP or Oracle. Then mobile phones came along, and suddenly we were building a hundred thousand mobile apps. There was huge demand. At IBM Consulting we have been doing this a lot. So we built the Yono app for the State Bank of India, which now has hundreds of millions of users.

 

When you think about agents and assistants — GenAI-based systems — they are essentially software, too. But instead of a hundred thousand, there could be a billion of them. And who will build them? Companies like ours — consulting firms.

 

So, I’m quite positive about the space and about consulting overall. But during this transition, the path will be bumpy.

 

Given reports that only about 5 per cent of companies have derived real value from AI, shouldn’t consultants have been more aggressive?

 

An MIT study said that 95 per cent of GenAI projects are failing to deliver returns. I actually believe that but I also think there’s a meaningful 5 per cent of companies that are truly succeeding.

 

About three years ago when ChatGPT popularised GenAI, we took that very seriously. Even though IBM has been in the AI business since 1957 and had GenAI capabilities, we weren’t the ones who brought a killer app like ChatGPT to market. So we decided that if we weren’t going to be the company to commercialise that breakthrough, we should at least be the most prolific users of GenAI within IBM.

 

Arvind (Arvind Krishna, chairman and CEO of IBM) launched that initiative. We broke IBM down into workflows that prioritised areas with the highest value and impact potential, and in the early days built 220 assistants across them — roughly four assistants per workflow. We now have about 3,000 such assistants and agents deployed. We achieved tremendous savings. In 2024, we saw $3.5 billion in savings. About two-thirds of that went right back into the business — we increased R&D spend from 9 per cent to 12 per cent. The company has added eight points of growth, and the stock price has doubled.

 

IBM Consulting used IBM as ‘client zero’, and today we are working with more than 100 clients, including several in India —Vodafone India is one example — where we are applying the same approach to implement AI in ways that work. That’s part of the 5 per cent that is delivering value.

 

I think consulting companies are really important in that regard. I think people had to try a bunch of things and have these things fail before recognizing that partnerships between enterprises and consulting firms are likely the best route to successful AI adoption.

 

We hear that when it comes to executing AI roadmaps, companies want to have all the control rather than work with consultants.

 

Yes, you’re right — initially, the mindset was ‘AI is core; we must do it ourselves’. But after people have experimented, succeeded in some areas and failed in others, they’re realising that, like with any new technology, you handle the parts you must own and you partner on the rest with those who have capabilities. That could mean working with consulting firms, hyperscalers, or software vendors. We’re starting to move past the ‘we have to do it all ourselves’ mentality.

 

You’ve said that services-as-software is the future. How is IBM implementing that model?

 

I have spent most of my career developing software. When I returned to IBM, I had never worked in consulting. One reason Arvind wanted me to take this role is because he and I had worked together building IBM’s software business many years ago. In his mind —and in mine — the consulting business, and many other businesses, will evolve toward a model that combines human workers with digital workers. A digital worker is essentially a small piece of software — an assistant or an agent.

 

We began building the capability to deploy these agents at scale. Today, we have about 3,000 unique agents and more than 100 projects with large clients where we are delivering using this human plus digital worker approach.

 

I was just in Bengaluru. At one client, we have about 1,500 people working this way; at another, about 1,000. For example, in one account, 835 of our consultants are working with 296 unique digital workers to get their work done.

 

It’s interesting how the term services-as-software has suddenly become popular, because we started building towards this two years ago — we just didn’t know how quickly it would arrive. Today, we have a platform inside IBM Consulting called IBM Consulting Advantage, where we host around 3,000 unique digital workers. At any given time, around 10,000 are running on the platform. They sit on top of multiple LLMs [large language models], supported by a common layer that lets us monitor and govern them.

 

Services-as-software is already here. And I actually believe IBM Consulting is meaningfully ahead in this area. Many others simply purchased Copilot or Gemini licences and distributed them across teams, hoping for productivity gains — which largely didn’t work. We, on the other hand, built a platform, integrated the LLMs, and developed domain-specific capabilities such as finance reconciliation.

 

Do you see the current softness in consulting getting resolved and by when?

 

It’s very hard to predict what will happen with the global economy. Much of it depends on factors unrelated to AI or technology —like trade policies and other macroeconomic forces. My assumption is that it will take some time before we return to a period of strong GDP growth. Until then, many economies will remain muted, which means continued budget constraints.

 

I think a second scenario is more likely to happen: Companies will need to prioritise their investments, especially in AI, over other areas. And that’s exactly what we’re seeing in consulting.

 

We are seeing tremendous growth in AI. It now represents about one-third of our business, by far the highest proportion among our competitors. But there are other areas where clients are de-investing. The net result is a muted consulting market overall, even as investment shifts strongly toward AI-related work.

 

How significant is India in your consulting business?

 

It’s very significant. India represents a large part of our workforce, and it’s an area where we’re investing heavily. We have a new hybrid Cloud center here. Some of our major areas of focus as a company include AI, cybersecurity, hybrid Cloud, and quantum. We have a cybersecurity center in Bengaluru that serves all of APAC, and we will be opening a quantum center later this year in India.

 

For IBM — both on the product and consulting sides — India is an extremely important location. We are also opening centres in the United States and in other countries with strong talent pools and innovation ecosystems. But India continues to stand out as a particularly critical hub for us.

 

As agentic AI becomes more ingrained in business, what impact do you see on people? IBM has also announced layoffs — how does this affect your business in India?

 

Let me address the second part first. Companies are always rebalancing their workforce, and for us, this is a very small percentage. There are areas where we are investing heavily and growing — AI, cybersecurity, hybrid Cloud, and quantum — and skills in those areas are expanding.

 

How this affects the workforce is a fascinating question, and one we started tackling as soon as we got into this. The first thing we did, we started training IBM-ers on AI skills. And we didn’t just stop internally — we created a global programme to train people in the broader community. We’ve now trained several millions of people globally.

 

I believe that, ultimately, there will be more jobs because the overall economy will expand. But there will be a shift in the types of jobs. Our message to employees has been: We will give you opportunities to learn, understand how to use AI, how to build with it, and how to use it as a tool. There will be a shift toward GenAI-enabled roles — whether it’s financial reconciliation or something entirely different — and there will also be categories of jobs that shrink.

 

So, net-net, I think there will be expansion. But that transition — the shift in job profiles — is something all of us will have to manage carefully.