ISG Provider Lens™ Mainframes - Services and Solutions - Application Modernization Services - Europe 2025

Europe’s mainframe landscape sees cautious investment, hybrid strategies and rising AI use

Across Europe, mainframes are crucial for sectors such as banking, government and manufacturing. Multiple studies estimate that European institutions process billions of transactions daily on mainframes. In the last 18 months, cost pressures, skill shortages and stricter regulations have sparked demand for services and software that optimize, modernize and secure legacy systems.

Hybrid strategies persist, balancing mainframe stability with cloud agility. In a recent mainframe client study by Kyndryl, most enterprises stated they now adopt a hybrid approach: 96 percent are migrating some portion of their workloads off the mainframe (on average 36 percent) while relying on them for mission-critical tasks. Meanwhile, 89 percent consider mainframes Extremely or Very Important, blending cloud scalability and mainframe reliability.

As enterprises integrate mainframes more closely with cloud and distributed platforms, these hybrid environments will demand new tools and processes — particularly to harness data across platforms for AI and analytics.

Europe’s slow or near-stagnant economic growth has made enterprises more cautious, favoring smaller, high-impact modernization projects that demonstrate clear ROI. At the same time, the ongoing need for competitiveness, compliance and AI-driven innovation propels continued investment in upgrading mainframes. As a result, modernization efforts in the region have shifted from large-scale overhauls to incremental, service-based approaches.

Generative AI (GenAI) has swiftly moved from being a PoC to an increasingly viable solution for mainframe modernization in Europe, assisting with code refactoring, documentation and skill transfer. While limitations in accuracy, regulatory complexity and resource requirements remain significant, continued advancements in domain-specific models and compliance frameworks promise to close existing gaps.

Skills shortage — a key issue in the European mainframe market

Mainframe clients in Europe have been struggling with the low availability of mainframe knowledge for years, and the issue persists. Eurostat data on workforce demographics shows that up to 30 percent of Europe’s IT professionals could retire or exit the workforce by 2030. Many of these professionals hold mainframe-specific skills such as COBOL, PL/I, IMS and CICS, making the retirement wave especially problematic for companies that still depend on these platforms.

Higher education institutions in the region are increasingly emphasizing cloud-native and webbased development, leaving mainframe-focused curricula less common. Surveys from various industry groups (including the European CIO
Association) indicate that less than 10 percent of recent IT graduates in Europe have exposure to mainframe environments, and fewer than 5 percent report proficiency in COBOL or assembler languages often used in legacy systems.

The fragmentation of the European continent into more than 30 countries with varied languages, partly different labor laws and diverse business cultures forces companies to seek talent on a limited local scale. Moreover, mainframe specialists require dedicated local language skills because teams in France, Germany, Spain or Italy may use local languages for system documentation or documentation of local regulations.

The result is a shallow talent pool for mainframe positions, even as the need for modernization expertise grows. Some of the largest companies in Europe face serious challenges in recruiting new talent with the security, compliance and 
mainframe-coded language skill sets needed to manage critical systems in an era of complex regulatory requirements. Therefore, leading regional banks and telecoms have set up internal mainframe academies, offering boot camps and labs to cross-train existing staff from other IT disciplines.

GenAI is driving change, with regulations hindering rapid (and more insecure) development

Over the past 18 months, the use of GenAI in mainframe environments has progressed from niche pilot tests to structured deployments. Early prototypes struggled with limited context and specialized legacy languages, yet the
technology demonstrated enough potential to attract broader interest. Initial trials focused on tasks such as  automated documentation and code translation, revealing that AI-driven systems can reduce the overhead of manual
analysis. Despite some accuracy challenges, these PoCs spurred further experimentation, highlighting how AI models could automate aspects of modernization projects and address enduring skill shortages in mainframe development.

Today, GenAI tools are being used to uncover hidden logic in COBOL and assembler code, refactor or annotate legacy applications and expedite documentation efforts.

In Europe, where banking, insurance and public administration often rely heavily on mainframes, these use cases  have begun advancing from pilot stages to production. Providers report faster discovery of dependencies, assistance in bridging knowledge gaps for younger developers and greater confidence in modernization timelines. However, the
technology still requires domain-specific tuning and human oversight; most organizations rely on subject matter experts to validate AI output and ensure it aligns with business requirements and compliance standards.

There are several limitations. Mainframe codebases are typically vast, customized and deeply embedded in critical workflows, rendering them less straightforward than modern systems. GenAI also risks generating hallucinations or convincing yet incorrect recommendations, making verification essential. Strict European data protection laws add complexity, as organizations must handle code and correlated data carefully to avoid breaching GDPR or sector-specific regulations.

In addition, AI-driven tools can be resourceintensive, whether they are run on-premises or via cloud services, and many companies remain cautious about how and where to train or deploy large models containing sensitive organizational logic.

GenAI in mainframe contexts is poised to expand across the development lifecycle. Expected improvements include the development of more precise domain-focused models to handle COBOL variants and legacy data structures, integrated compliance features that log and audit AI use, and deeper ties to DevOps pipelines for automated testing. Many providers are working on capabilities supporting hybrid cloud architectures, where modern and legacy systems interconnect seamlessly. Enhanced low-code integrations may also emerge, allowing non-experts to leverage mainframe data while relying on AI-driven code insights for complex backend processes.

The discussion about the limitations that Europe’s regulatory environment puts on AI development and how  regulations can be adapted to fuel development without loosening its protection mechanisms is in full swing.

Its result will play a unique role in shaping GenAI adoption for mainframe code analysis. Today, strict data protection rules and local sovereignty requirements necessitate either robust on-premises solutions or carefully managed  partnerships with AI providers. Large banks and government agencies have already pushed for specialized security  and auditing features, influencing global providers to accommodate European privacy mandates. These constraints sometimes slow adoption but ultimately contribute to more reliable, well-documented implementations, setting a
precedent for international deployments that must prioritize security, compliance and data ethics.

The necessity for innovation is stronger than pure cost control

Europe’s mixed economic outlook, with several leading economies experiencing nearstagnation or minimal growth, has created a dual effect on mainframe modernization. On the one hand, tighter budgets and costconscious strategies have prompted some enterprises to proceed more cautiously with large-scale investments. Decision-makers often look to defer or phase out capital-intensive modernization projects, focusing instead on smaller, quicker improvements with clear ROI. This approach has led many companies to scrutinize every outlay more closely, seeking ways to justify  projects through measurable gains in efficiency and resilience.

On the other hand, the drive to remain competitive — especially against global competitors and in the face of new digital mandates — continues to power modernization initiatives. Even in a slow-growth environment, organizations cannot afford to stagnate core infrastructure to maintain or improve market share. Compliance requirements, security
threats and the surge in AI-driven innovations provide additional impetus as outdated or inflexible systems become bottlenecks to agile service delivery. In regulated sectors such as banking, insurance and government, modernization is often viewed as essential to meeting evolving standards and customer expectations, rather than a discretionary expense.

As a result, the trend toward modernization has not halted; it has merely shifted. Many enterprises now pursue incremental modernization — updating specific workloads, integrating targeted AI solutions and employing hybrid models that blend mainframe stability with cloud flexibility. Outsourced operations services, mainframe-as-a-service models and phased modernization approaches also help reconcile cost constraints with the push for innovation. Consequently, despite economic headwinds, the need to stay competitive and comply with rigorous regulations  ensures that mainframe modernization remains a priority in Europe’s business landscape.

Mainframe Optimization Services

European enterprises rely on mainframe optimization to curb costs, adapt to inflation and improve resource utilization. In light of economic uncertainty, many European enterprises are focusing on cost efficiency, leading to a noticeable uptick in demand for specialized teams that can tune mainframe environments (e.g., CPU utilization, workload 
management and capacity planning).

Over the last 18 months, AI-powered diagnostics have highlighted performance bottlenecks, driving down CPU  overhead. Heightened regulatory pressures (e.g., GDPR and NIS2) now shape optimization priorities and have pushed optimization services to include data protection and audit-readiness assessments. By combining capacity planning with cost oversight, these services extend mainframe life, streamline processes and unify monitoring across hybrid environments.

Mainframe Application Modernization Services

A hybrid approach to modernization remains prevalent across Europe. Rather than complete migration off the  mainframe, many enterprises favor strategies that retain mission-critical workloads on mainframe infrastructure while
integrating cloud services for front-end or analytics functions. As COBOL experts retire, service providers have become increasingly consultative, offering training and code analysis to address skills shortages. GenAI is also being applied for automated code refactoring and documentation. In the past 18 months, AI-driven analysis has eased skill gaps by automating COBOL translation and revealing dependencies.

In BFSI, as well as national or regional government agencies, application modernization services are in particularly high demand to meet compliance and digital transformation mandates, with a drive to accelerate modernization initiatives. This shift has spurred providers to adopt Agile and DevOps practices that shorten project cycles and reduce risks.

Mainframe Operation Services

As internal mainframe teams shrink, many European enterprises increasingly turn to outsourced/managed services to maintain mission-critical systems. This approach frees them from routine tasks such as monitoring, patching and incident response while letting staff focus on innovation.

Beyond basic uptime and performance SLAs, buyers now expect advanced security controls (such as encryption, real-time threat detection and compliance reporting), forcing operation service providers to expand their offerings. Therefore, recent demand centers around end-to-end observability — one console or dashboard that can track  mainframe performance along with distributed and cloud services.

More operators are introducing AIOps to meet these demands, reduce manual intervention and speed up incident resolution. Although adoption is still in the early stages for many, the trend is gaining ground in large-scale environments looking to streamline support. ISG expects some significant development in this area in the next 18 months.

Mainframe as a Service (MFaaS)

While the MFaaS model has gained more visibility in the last 18 months, adoption in Europe remains measured, and the number of providers offering MFaaS at scale remains low. MFaaS is pitched as a way to convert mainframe overhead into a more OpEx-centric model. However, understanding TCO can still be tricky, leading some organizations to pilot MFaaS solutions before fully committing.

Organizations in highly regulated industries often cite data sovereignty concerns and specific compliance obligations that require either on-premises hosting or stringent contractual agreements that only a few providers can meet.

MFaaS can help scale up or down without large capital outlays for companies with seasonal or variable workloads. This advantage resonates with midsize and small enterprise clients that need mainframe capabilities but lack the budget or the talent to manage everything internally. The move to MFaaS also creates a good opportunity to bring the  mainframe environment up to date with the latest patches and upgrades, which might have been delayed due to the rising lack of in-house skills. Early adopters often pair MFaaS for certain test or development environments with on-premises mainframes for production. This strategy aligns with a broader hybrid IT philosophy.

Mainframe Application Modernization Software

Over the past 18 months, the availability of AIenhanced code analysis and conversion tools has increased. These tools aim to accelerate modernization projects by generating reports on code dependencies, business logic and application flows. Some solutions enable lowcode or no-code development on mainframe assets, letting non-mainframe specialists create or integrate new services without delving into COBOL or assembler.

As more European enterprises adopt modern CI/CD practices, the demand for software that easily plugs into  enterprise DevOps toolchains has surged, allowing consistent build, test and release cycles across both cloud and
mainframe components. Many modernization software suites offer built-in security checks, encryption libraries and compliance audit features to align with European data protection and privacy regulations. As many providers of mainframe application modernization software are headquartered outside the EU and might have limited local  workforce capacities, they rely heavily on their partnerships with multinational service providers to create a localized offering and ensure the availability of regulatory skills and processes.

Access to the full report requires a subscription to ISG Research. Please contact us for subscription inquiries.