Where’s IBM in the AI craze?

Like much of the tech industry, IBM is racing to perfect agentic AI, but it’s not drawing the same buzz as rivals like OpenAI and Google. Some analysts believe it’s more of a PR than a product problem.

The market has seen IBM “significantly” lag behind Microsoft in AI and cloud growth because of the focus on tailored solutions rather than scalable cloud services for the public, according to Seeking Alpha’s 2025 reporting.

Making a case. But Matt Lyteson, IBM’s CIO and VP of technology platform transformation, said IBM has been a leader in AI for years. His goal is to capitalize on IBM’s technology and scale it for the enterprise.

“We very quickly asked ourselves: ‘What does agentic AI look like at the enterprise scale, and what are we going to focus on?’” Lyteson said. “When we think about the enterprise context, the real value is doing these end-to-end workflows.”

Lyteson believes it’s “great” that companies can garner attention with AI consumer experiences, but enterprises have a totally different set of questions and concerns, including how to ensure agentic identity management, equality of interactions with team members, or safety with interactions.

“Those are all the things that IBM is addressing head on,” Lyteson said. “Quite frankly, the market’s ignoring that.”

What’s the issue? While Microsoft has Copilot and Google has Gemini, IBM isn’t driving AI adoption via a single product or platform, which can make it hard to determine its place in the AI conversation, according to Rick Villars, group vice president of worldwide research at IDC.

“What IBM does—how we assess their impact—is [that] their strength as an organization broadly is with larger enterprises globally,” Villars said. “They’re at the core of many businesses.”

But the broader market, Villars said, doesn’t have high visibility into IBM’s solutions or deployment, and therefore the company is judged by how effectively it helps customers navigate through the AI craze.

“It’s a complex story that they have, and so they have to deal with that…Because as they’ve pivoted as a company, they’re in midstream of story development,” Villars said. “They have a very loyal customer base of people who want to read the story and want to see the drafts and everything else, but go beyond that and to build that presence.”

Issues with Watson. IBM’s AI tool Watson, which could be seen as synonymous with Microsoft’s Copilot, was spun into existence almost 20 years ago as a platform for competing on Jeopardy!. In 2020, IBM released a beta version of Watson Assistant that acted as a chatbot, combining machine learning, transfer learning, and deep learning techniques onto a single platform.

IBM then released Watson Health in 2015, designed to help with cancer research and other healthcare issues, which subsequently ran into significant challenges and was sold off parts to a private equity firm in 2022.

Now, IBM has Watsonx, an umbrella term for the company’s portfolio of AI products. As of December 2025, Watsonx can monitor agents at runtime, giving IT pros insight into how their agentic AI is performing and how users are interacting with it.

The old Watson, Villars said, is not the same as Watsonx: “While they’re very different, it does affect how people see what IBM was doing or is doing, so it’s something that they have to work hard to be very clear about what they’re explaining and what it is. I’d say it’s a challenge that they do face.”

What we can expect to come. There’s one area where IT professionals can expect to see IBM go more in-depth: data governance,which remains a key part of agentic AI.

Villars said this is more about data in motion, as a result of agentic AI focusing on “event-driven activity” (such as when a software or smartphone company tracks how people use a particular app).

“IBM has an amazing event stream…but they didn’t have a very AI-connected way of making that event information consumable,” Villars said. “I think you’re going to see announcements around how they can.”