IBM Targets Mainframe Modernization With Z Atlas AI Package

IBM has announced Z Atlas, an AI-powered package designed to map a client's entire z/OS stack and eliminate the leading cause of unplanned mainframe outages: changes that are not proven to be safe before they reach production. The announcement, detailed in a blog post by Skyla Loomis, general manager of IBM Z Software, positions AI as the mechanism for keeping workloads on the mainframe rather than migrating them off it.

The product exists today only as a "statement of direction," meaning IBM has described its intended capabilities but has not shipped them. Customers can already sign up for a preview. The stakes for IBM are considerable: a lineup of competitors, including AWS, Google Cloud, Microsoft Azure, OpenText/Micro Focus, and Accenture, actively markets services to move customers off Big Iron and onto cloud or distributed platforms.

Why Does Mainframe Change Management Need AI?

Loomis framed the problem in terms of process failure and demographic pressure. Many mainframe teams still rely on manual analysis, fragmented tooling, and institutional knowledge to understand dependencies and evaluate risk across teams — a combination she described as slow and ineffective.

"As experienced professionals retire and modernization initiatives accelerate, the siloed knowledge problem is magnified with mainframe teams having trouble measuring the full impact of their changes before they reach production," Loomis stated.

Z Atlas is designed to close that gap. According to Loomis, the package will map a client's entire z/OS stack across infrastructure, middleware, and applications, then use that system intelligence to reduce the manual effort of tasks such as:

• Patch management
• Regulatory response
• Knowledge sharing across every team, from systems programmers to application developers

The core idea is to map how a change impacts the z/OS stack by testing and validating it before deployment. Loomis offered a concrete use case in her blog post.

"For example, z/OS system programmers can use this solution to evaluate how a middleware upgrade might affect dependent applications, identify related infrastructure and operational processes that may require validation, determine which systems should be included in a test plan, or coordinate validation activities across application, infrastructure, and operations teams," Loomis wrote.

What Technology Does Z Atlas Use?

The technical foundation, as Loomis described it, rests on five capabilities working in concert. "IBM Z Atlas' system intelligence is built on a foundation of automated discovery, topology modeling, dependency analysis, natural-language interactions, and AI-powered workflows to create impact assessments and change evidence," she stated. "Together, these capabilities are intended to create a continuously updated understanding of the IBM Z environment and the relationships between its components."

Topology modeling (a live map of how system components connect) and dependency analysis (identifying which applications rely on which middleware and infrastructure) address the fragmentation problem directly. Natural-language interaction would allow teams to query their environment without specialized tooling knowledge.

The broader vision extends beyond change validation. Loomis said IBM intends Z Atlas to become "the shared intelligence layer that will enable Z teams to understand, manage, govern, and quickly optimize their IBM Z environments through natural-language experiences and AI-assisted workflows."

How Does the Market Context Support IBM's Bet?

Z Atlas arrives as independent analysts challenge the mainframe-exit market that IBM is defending against. A recent Gartner report warned that poor mainframe migration decisions threaten business and operational continuity.

"Organizations that pursue 'seemingly magical' AI-driven exit strategies, rather than taking a platform-smart approach to align workloads with the right environments, risk introducing significant technical debt and exposing the enterprise to critical failures," stated Alessandro Galimberti, vice president analyst at Gartner, in the report.

Gartner's quantitative forecast is stark. The firm predicts that by 2030, 75% of vendors operating in the mainframe exit market will pivot their business models or cease operations, as market expectations reset and demand for one-size-fits-all migration solutions declines.

"As a result, the mainframe exit market is at a pivotal point. Sustained investments from IBM, along with independent software vendors such as 21CS, BMC, Broadcom, and Rocket Software, and managed service providers like DXC, GTSG, and Kyndryl, further reinforce the mainframe's position as a modern platform for strategic investment," the Gartner report stated.

Galimberti argued that the right approach depends on the customer, not the fashion of the moment. "The right mainframe strategy depends heavily on the profile and complexity of an organization's environment," Galimberti said. "For many mainframe customers, GenAI can be more effectively used to enable modernization in place rather than accelerate migration off the platform."

That argument aligns precisely with what Z Atlas does: it applies generative AI and agentic workflows to modernization in place, on the mainframe itself, rather than to migration off it.

What Do IBM's Own Numbers Show?

IBM's financial leadership supplied the commercial evidence for the strategy during the company's July earnings call. James Kavanaugh, IBM senior vice president and CFO, reported that AI is driving incremental capacity growth and new workloads as clients look to run AI closer to their most sensitive data.

"We are seeing strong early adoption of our AI innovations with nearly 50% of z17 customers investing in AI capabilities with Spyre AI accelerator, and clients deploying Watson X Code Assistant for Z are growing MIPS capacity three times faster than those who are not," Kavanaugh said.

He also quantified the cost argument for staying on the platform. "Depending on the size and complexity of workloads, clients can realize a 2 to 15x total cost of ownership benefit versus moving these workloads off the platform, reinforcing why the platform remains central to their operations and positioning us to capture additional value as AI workloads grow," Kavanaugh continued.

The numbers matter for the Z Atlas business case. If AI-assisted tools such as Watson X Code Assistant for Z (IBM's code-generation and modernization assistant for mainframe developers) correlate with three-times-faster MIPS capacity growth, extending AI intelligence into change management and system-wide dependency mapping targets the same adoption flywheel from a different angle.

What Happens Next?

Z Atlas remains a preview-stage commitment rather than a shipping product. Its success will depend on whether automated discovery and dependency analysis can genuinely replace the institutional knowledge walking out the door with retiring mainframe veterans. With Gartner forecasting that three-quarters of mainframe-exit vendors will pivot or fold by 2030, and IBM claiming a 2-to-15-times TCO advantage for staying on Z, the battle over where AI-era workloads run is now being fought with AI tools on both sides.

By Marcus Bennett