Gartner warns most ai driven mainframe migrations will fail

Thursday 16 April 2026 - 10:20

By: Dakir Madiha

Gartner warns most ai driven mainframe migrations will fail

Gartner has warned that a large majority of enterprises relying on generative AI to migrate away from mainframe systems are likely to fail. In a recent report, the firm estimates that more than 70 percent of such projects launched in 2026 will not meet their objectives, citing an overestimation of what AI tools can deliver in complex legacy environments.

The report argues that generative AI can help identify and document technical debt within mainframe systems, but falls short when it comes to fully automating code conversion and migration. These systems often support critical workloads with high performance and throughput requirements that are difficult to replicate outside mainframe environments. As a result, organizations risk losing key capabilities during poorly executed transitions.

Analysts highlight the scale and interconnected nature of enterprise data as a core barrier. Many large organizations operate systems built over decades, with tightly coupled processes and dependencies. According to the report, this level of complexity makes full scale migration both technically challenging and financially prohibitive in most cases.

The study also points to market dynamics driving unrealistic expectations. Investor pressure has pushed software vendors to position AI as a universal solution, encouraging the development of migration tools that may not be suited to real world enterprise needs. At the same time, companies face genuine concerns about aging mainframe expertise and growing technical debt, which increases the appeal of AI driven solutions.

The warning follows earlier developments involving Anthropic, which promoted its Claude Code tool as a way to modernize COBOL systems. That announcement triggered a sharp market reaction, including a significant drop in the stock of IBM, a long time leader in mainframe infrastructure. Gartner’s analysis challenges this narrative, emphasizing that mainframes remain essential for certain mission critical applications.

Instead of pursuing full migration, the firm advises organizations to modernize existing mainframe systems incrementally. The report stresses that failed migration efforts can have severe consequences, including operational disruption and business continuity risks. For many enterprises, maintaining and evolving current systems may offer a more reliable path than attempting a complete transition driven by immature AI capabilities.