The surge in artificial intelligence (AI) capabilities is reminiscent of past technological booms, filled with exciting demonstrations, eye-catching market valuations, and a slew of tools boasting instant solutions to myriad problems. Global spending on AI is predicted to exceed $2.5 trillion this year. However, as the initial excitement dies down and organizations begin to seek real returns on their investments—covering aspects like uptime, security, costs, and compliance—the critical question will shift from “What tools did you choose?” to “Have you built something sustainable?”
Tools alone are not solutions; they are merely instruments to assist in achieving goals. At IBM, we don’t just offer tools—we are committed to building lasting solutions. This commitment is evident in our approach to mainframe modernization, which has seen remarkable platform growth of over four times in the last decade. Creating enduring value involves more than selecting the right tools; it requires the right platform and a strategic approach.
Modernization Starts with the Right Platform
Modernization transcends simple COBOL translation or being a one-off event. Translation addresses immediate needs, but modernization represents an ongoing journey. It involves continuously leveraging new tools while safeguarding past investments and securing operational environments. Additionally, modernization seeks to simplify an increasingly complex application landscape, characterized by vast interdependencies and growing workloads. To navigate this complexity, tools like watsonx Code Assistant for Z and watsonx Assistant for Z are essential, enabling organizations to preserve institutional knowledge and create more intuitive interfaces for developers and users alike.
Most importantly, modernization encompasses the entire technology stack—not just individual software components. From chip architecture to operating systems, middleware, security frameworks, automation, and AI services, value is derived from how well each layer works together. IBM Z is specifically designed to handle transaction processing, delivering unparalleled efficiency, scalability, performance, and resilience. Choosing any other platform invariably involves compromises.
We enhance every layer of the stack to achieve optimal outcomes. Our IBM Z systems are capable of processing tens of billions of encrypted transactions per day, delivering AI inference at remarkable speeds, and achieving industry-leading uptime while maintaining security standards. Transitioning code alone cannot replicate these attributes; it is the design of the platform that matters, not just the programming language.
This commitment to modernization is precisely why IBM pioneered the concept: to simplify access to mainframe applications, fostering innovation and agility while managing risks and optimizing costs. Our ongoing efforts in mainframe application modernization empower clients to derive greater value without abandoning systems that excel at their jobs. For instance, we introduced the first generative AI-powered code assistant, watsonx Code Assistant for Z, in 2023, specifically targeting IBM Z.
Since its launch, countless clients have adopted this technology to embark on modernization initiatives. By leveraging AI and analytics, organizations can maximize the value of their data without the costs and risks associated with data migration from the mainframe. A notable example is Swiss Mobilar, which utilized SQL Data Insights software on IBM Z mainframes to achieve near-real-time AI inferences with 94% prediction accuracy, enhancing insurance quotes closure by 7% within just six months.
As we enter our fourth year of utilizing AI accelerators on IBM Z for transaction workloads and our third year for application development, we emphasize that the focus should not be on abandoning institutional knowledge. Instead, organizations should seek to understand, enhance, refactor, and connect that knowledge with the AI capabilities that will drive growth in the coming decade.
The Right Approach: Mainframe Modernization Goes Beyond Vibe Coding
Many enterprise IT teams subscribe to the 80/20 rule: approximately 80% of application code operates effectively. The remaining 20% is the key to unlocking business value. Updating that critical code while keeping it on-premises is faster, more cost-effective, and less risky than rewriting everything for an alternative platform that wasn’t designed for high-volume transaction processing. Discussions with numerous clients have revealed that they do not seek language migration; they want better access to essential capabilities, such as open APIs, hybrid cloud integrations, and secure data sharing for model development—all without compromising on performance, quality of service, or resilience.
On the mainframe, AI acts as a catalyst rather than a compromise. It helps bridge the skills gap by enabling developers—regardless of their programming backgrounds—to understand, explain, and enhance mission-critical code. This technology accelerates the adoption of modern, git-based DevOps practices on IBM Z, facilitating seamless CI/CD workflows and genuine continuous delivery. Moreover, it streamlines the challenging aspects of modernization including deep code analysis, dependency mapping, performance optimization, and risk-free refactoring.
Success in this arena demands more than mere models relying on guesswork. Enterprise applications can comprise millions of lines of code and thousands of artifacts. IBM’s unique approach combines large language models (LLMs) with essential contextual information—thorough static analysis across the entire application landscape and dynamic runtime data from z/OS that illuminates actual execution paths and performance bottlenecks. This comprehensive insight enriches our AI software, leading to more precise outcomes. Nearly half of our z17 clients have already invested in new AI capabilities, with the adoption of generative AI software on IBM Z rising by 200% year over year. Such advancements make it possible to modernize without sacrificing resilience or security.
Clients are actively validating this approach. The Royal Bank of Canada is using AI to enhance understanding and blueprint their modernization across complex systems. The National Organization for Social Insurance (NOSI) significantly reduced the time needed to analyze redundant COBOL routines, allowing teams to prioritize more impactful changes. ANZ Bank employs modern DevOps on IBM Z to accelerate project delivery while minimizing manual intervention. These endeavors go beyond surface-level migrations; they reflect quantifiable gains in speed, safety, and value on the platform.
The Future Belongs to Builders
Another misconception worth dispelling is that a solely SaaS-based solution can effectively “modernize the mainframe” from a distance. Given the on-premise dependencies, compliance requirements, and sensitivity surrounding core transactions, this idea fails to withstand scrutiny in terms of governance and risk management. The route to sustainable growth is inherently hybrid. Essential transaction processing should remain on the platforms designed to handle them, while cloud-native capabilities should be integrated where they add value, utilizing AI to enhance connectivity between both environments more efficiently and securely.
Over 90% of the world’s credit card transactions are processed through IBM Z for good reason. While other sectors explore GPUs and alternative platforms, IBM Z continues to be the best solution available for transaction processing—from the silicon level to the operating system and middleware. Attempting to replicate its capabilities elsewhere will only lead to increased costs, compromised performance, and reduced resilience and security.
As the excitement from the AI boom wanes, true prosperity will belong to those who recognize the difference between tools and real outcomes. They will partner with organizations capable of modernizing the critical 20% of their code, maintaining performance guarantees, and seamlessly integrating AI into the infrastructures best suited for their needs.
This is where IBM stands: not merely as a purveyor of tools, but as a builder equipped with the right platform and strategy for sustainable enterprise IT. In this new age of AI, the strengths of the world’s foremost transactional platform will only continue to amplify.
