84% of Mainframe Users to Rely on COBOL Despite AI Conversion Limits
The persistence of mainframe reliance signals enduring legacy challenges, unlike rapid shifts seen in cloud adoption.
Key Points
- 1Mainframes remain vital, reflecting legacy reliance and modernization struggles.
- 2Limited generative AI capabilities highlight technical hurdles in mainframe conversion.
- 3France's Bercy aims to enhance AI policy, boosting national tech capabilities.
What Changed
A recent IDC MarketScape report indicates that 84% of mainframe users intend to rely on mainframe technology for the majority of their workloads over the next five years. This decision occurs despite ongoing challenges in converting COBOL code—often comprising millions of lines—into more modern languages like Java or Python using generative AI. The persistence of mainframes highlights a continued legacy system dependence due to their deep integration and critical functionality.
Strategic Implications
The limited success of generative AI in converting COBOL presents both a challenge and an opportunity. Companies relying on these systems must weigh the risk of maintaining aging technology against costly modernization efforts. Bercy, France's Ministry of Economy and Finance, has taken a proactive step by establishing a direction to enhance AI and digital strategies, which could bolster France's competitiveness in technology sectors. This move underscores a strategic pivot to solidify national sovereignty in AI capabilities.
What Happens Next
As AI capabilities slowly evolve, we can anticipate gradual improvements in AI’s ability to assist with code conversion. However, immediate reliance on fully autonomous conversion tools remains unrealistic. Instead, hybrid approaches that combine automated and manual processes may gain traction. France is likely to leverage its new AI direction to craft policies that support innovation while managing the transitional risks associated with legacy systems.
Second-Order Effects
The wider tech ecosystem might experience ripple effects, especially in workforce dynamics. As experienced COBOL developers retire, there may be increased demand for training programs to equip new developers with skills to manage legacy systems. Additionally, firms offering AI tools will seek partnerships with enterprises needing tailored solutions, targeting markets with heavy mainframe reliance.
Free Daily Briefing
Top AI intelligence stories delivered each morning.