Insights at a Glance
Generative AI will impact three separate domains: code understanding, code conversion/completion, and IT system operations (e.g., automation, subsystem administration, etc.). The maturity and effectiveness vary across these three areas.
Many organizations make mainframe strategy decisions based on emotion and hype rather than basing their decisions on facts and not distinguishing between mainframe migration and modernization, which vendors often conflate. Recent GenAI advancements have only added to the confusion.
The specific profile your organization (large, medium, or small) falls into directly affects the appropriate strategies for modernizing or migrating your mainframe environment. For many large organizations AI is more of an enabler for modernization in-place than migration off the mainframe.
Avoid viewing mainframe migration as just finding a tool that can convert COBOL-based applications. Weigh the risk profile of the mainframe versus alternative platforms, distributed or cloud systems, noting that these platforms still cannot provide out of the box the same level of resiliency, security, high availability and transactional integrity offered by the IBM mainframe.
Strategic Planning Assumptions
More than 70% of mainframe exit projects initiated in 2026 will fail to produce the intended benefits due to an overestimation of generative AI tooling capabilities.
By 2030, 75% of vendors operating in the “mainframe exit” market will either pivot their business models or cease to exist.
Issue
An organization’s strategic roadmap for the contemporary mainframe landscape must be determined by three distinct environmental profiles, which are categorized by the size of the environment in MIPS.
The mainframe is still the leading platform for certain mission-critical applications, even with the ongoing drive toward cloud-native architectures. This is due to its unique, built-in capabilities for unmatched resiliency, security, high availability, and transactional integrity. A primary driver for its continued dominance is the ability to exceed “five 9s” (99.999%) of availability, a benchmark of reliability and uptime that remains structurally difficult to replicate within the distributed nature of public cloud environments. Other platforms might try to offer similar outcomes, but these do not come out of the box and require careful planning and implementation within the application logic.
Because the mainframe is often the only platform capable of providing these capabilities out of the box, decades of transaction history have accumulated on-platform, creating significant data and AI gravity. For most large-scale enterprises, the sheer volume and interconnected complexity of this data make wholesale migration a physical and financial impossibility. Consequently, the strategic focus for many of these large-scale deployments has shifted from moving data to the AI, to bringing generative AI and analytical capabilities directly to the mainframe (e.g., leveraging the AI capabilities on the mainframe (see “Is the IBM Mainframe Ready for AI?).
The drive to abandon the mainframe is diminishing. Customers are increasingly recognizing the near-impossibility of a mainframe exit at an acceptable cost and risk, leading them to give up on the long-held hope for a perfect tool to achieve this migration. While generative AI is currently transforming the “discovery” phase — aiding teams in mapping extensive technical debt — and improving operational support through capabilities such as specialized CICS and JES agents for an aging workforce, it still has significant limitations when it comes to the automated conversion and migration of legacy code. It also does not account for the unique capabilities that the mainframe offers (e.g., ensuring that the same performance and throughput is achieved after the migration).
Customers recognize the important value proposition of IBM’s continued investment to keep the platform modern and fit for purpose. The platform offers out-of-the-box capabilities for mission-critical applications and immense backward compatibility, allowing code written over 50 years ago to run without modification. These features significantly reduce risk and long-term TCO.
Impact
IBM mainframes remain the functional backbone of the global economy, supporting the most critical workloads in banking, insurance, healthcare, and government. However, for the modern enterprise IT strategist, navigating this domain is increasingly treacherous due to several converging variables:
The gap between the “marketing promise” of generative AI and its actual capabilities in code transformation.
Aggressive investor demand for AI capabilities as the sole indicator of a vendor’s long-term health forcing vendors to deploy AI even where unnecessary.
The high criticality and “too-big-to-fail” nature of core mainframe applications.
An aging workforce and the accelerating loss of specialized institutional knowledge.
The stakes of a miscalculation are immense. Poor decision making regarding migration is not merely a budgetary overage; it is a threat to business and operational continuity. Falling for “seemingly magical solution” migration promises while ignoring a platform-smart approach (i.e., diligently evaluating your workloads and choosing the best platform for the relevant work) leads to massive technical debt and critical enterprise risk.
Implications
Actionable Guidance by Environment Size
Your strategic direction is primarily determined by the size of your environment measured in millions of instructions per second (MIPS). While this is not a definitive indicator, it provides a useful measure of the complexity of the environment and, as MIPS are tied with spending, also your organization’s capabilities to invest.
Large Mainframe Environments (25,000+ MIPS)
For large mainframe environments, the scale of operations makes a total exit nearly impossible, nor justifiable by an increase in value or business outcomes. Therefore, the typical strategy is one of relentless optimization. A “platform-smart” strategy is the foundation for this profile, requiring organizations to identify and rehost workloads to the most appropriate platform over the long term, maximizing the advantages offered by that platform.
The mainframe is best suited for applications that support stable business processes, have a long-expected life span and demand high security and high-volume transactional integrity. Conversely, the cloud and distributed environments are more appropriate for applications that change rapidly, require significant burst-in scalability and are geographically dispersed.
After defining their “platform-smart” remediation roadmap, organizations should pursue the following primary options:
Dedicated relationship: Keep a direct relationship with IBM, and purchase directly from IBM and ISVs related to their mainframe estate, without using intermediaries such as hardware and software resellers or mainframe-as-a-service (MFaaS) providers.
MIPS optimization and predatory ISV exit: These are usually quick-win activities to improve platform’s cost efficiencies, freeing up investment power for additional innovation within the platform. They are often achieved in partnership with experienced system integrators such as DXC Technology, GTSG and Kyndryl. See Reduce IBM Mainframe Technical Debt Without Risky Exit Strategies for further details. MIPS optimization: Optimization projects aim to identify and target top MIPS consumers to reduce consumption, offload tasks to specialty processors, and lower or contain ISV and hardware costs.
Predatory ISV exit: Migrating to a different ISV can reduce software spending with a low-to-moderate risk and offer a fast ROI.
Mainframe modernization: With the resources freed up through predatory ISV exit and MIPS optimization projects, invest in modernizing your mainframe environment by enabling a modern developer experience, implement DevOps, enable API communications to include your mainframe in your hybrid estate, invest in automation, and incorporate AI in your mainframe. See How to Modernize the Mainframe Environment for further details. Mainframe platform architect team: Build a mainframe architectural team responsible for maintaining the platform’s relevance, ensuring technical governance and promoting awareness of modern mainframe technologies. Heads of I&O Must Embrace Modern IBM Mainframe Skills and Technologies describe which are the primary tasks for a mainframe architectural team within your organization.
Medium Mainframe Environments (5,000 MIPS to 25,000 MIPS)
Medium environments constitute the largest market segment. Their positioning presents the most complex decisions, as they must navigate between strategies focused on “optimize” and those centered on “exit.” The projected adoption is based on extensive Gartner interactions with both customers and vendors in the market (see Table 1).
Strategic path | Projected adoption (through 2030) | Description |
Do nothing | 40% | Maintaining the status quo and doing small tactical mainframe modernization projects. |
Mirroring large mainframe environments
| 30% | Adopting the same hosting, optimization, modernization, and approaches used by 25K+ MIPS environments. |
Rearchitect, rebuild, and/or replace | 15% | Significant, high-risk transformation efforts to move workloads to cloud-native architectures. Often the ending is suboptimal outcomes. |
Replatform/MFaaS | 15% | Using solutions to move workloads to x86 while maintaining code logic, or full outsourcing with the MFaaS consumption model to “set and forget” their mainframe estate. |
|
Source: Gartner (April 2026)
Small Mainframe Environments (Less than 5,000 MIPS)
Small mainframe environments struggle to achieve the economies of scale necessary to justify the significant overhead associated with mainframe migration or modernization. Yet, for many of these organizations, a positive return on investment (ROI) for a complete migration is often difficult.
This projected adoption is currently divided: 50% is adopting a “minimalist” strategy, either through replatforming with tooling or opting for “set and forget” MFaaS hosting. The remaining 50% employ a set of tactical opportunistic improvements to enhance their long-term TCO and their technical fitness.
Strategic path | Projected adoption (through 2030) | Description |
Replatforming or MFaaS | 50% | In this segment, deployments are moving toward replatforming or MFaaS. Replatforming solutions or MFaaS allows these organizations to exit the expensive hardware cycle while preserving the business logic of their mainframe applications.
|
Opportunistic improvements | 50% | MFaaS and innovate: Shifting to a consumption-based cloud model for mainframe hardware and implementing quick-wins mainframe modernization activities. ISV swaps: Replacing expensive independent software vendor tools with lower-cost alternatives. Specific workload removal: Moving peripheral applications off the platform while the core remains. Status quo: Doing nothing because the cost of the move exceeds the potential savings of the exit.
|
|
Source: Gartner (April 2026)
Summary of Environmental Profiles
Table 3 shows mainframe environmental profiles.
Environment size | MIPS range | Primary strategic paths |
Large | 25K+ MIPS
| Platform-smart, dedicated hosting, continuous modernization, predatory ISV exit and MIPS optimization. |
Medium | 5K-25K MIPS | Status quo (40%), mirroring large shops (30%), rearchitect, rebuild, and/or replace (15%), replatform/MFaaS (15%) |
Small | < 5K MIPS | Replatforming (50%) or tactical improvements (50%). |
|
Source: Gartner (April 2026)
This research is based on over 2,500 client interactions over the past 3 years.