According to Gartner, more than 70 percent of mainframe exit projects will fail by 2026. Furthermore, the analyst firm expects that 75 percent of the vendors involved will disappear by 2030. This pessimistic outlook is based on the limitations of generative AI when migrating legacy code.
Gartner recently published a report outlining the challenges faced by companies using AI to migrate legacy code from mainframes to alternative platforms. According to the firm, expectations surrounding AI-driven migrations are too high, resulting in many projects failing to meet their objectives. Gartner warns that over-reliance on AI tools and investor pressure on vendors to promote AI solutions are contributing to the problem.
Complexity and technical debt
One of the biggest challenges in migrating mainframes is the complexity and volume of data residing on these systems. Mainframes are often responsible for mission-critical applications and contain data accumulated over decades. Gartner states that switching to alternative systems often proves physically and financially impossible. While generative AI can help identify technical debt, it falls short in the actual migration and conversion of code.
Analysts emphasize that AI cannot guarantee that the unique performance and throughput capacity of mainframes will be maintained after migration. This means companies risk operational disruptions and significant technical debt if they rely too heavily on AI for their migration projects.
Caution and alternative strategies
Gartner advises companies to be cautious of AI’s promises regarding mainframe exits and instead focus on ways to improve their existing systems. The firm notes that many clients realize a full exit is often not feasible without unacceptable costs and risks. It is essential for companies to evaluate their workloads and choose the best platform for their specific needs, rather than relying on “magical” AI solutions.
Conclusion
Gartner’s outlook will please companies like IBM, which still have great confidence in mainframes for certain critical applications. Despite the pressure to move to cloud-native architectures, mainframes remain a reliable choice for many businesses. Gartner emphasizes that a well-considered approach is essential to managing the risks and costs of migration projects.