In an era defined by AI hype and cloud-first declarations, the mainframe environment, in contrast, appears quiet and refined. Organizations running IBM Z continue to anchor revenue, enforce governance, and modernize on their own timetable. Nothing is happening reactively or nostalgically. Movement continues, but only strategically.
What emerges from the 2026 Arcat Mainframe User Survey’s findings is discipline.
Five signals stand out from the data. Read them, then download the full report.
Five key findings from the 2026 data
1) You Can’t Call it Legacy if it Runs the Money.
The concentration of revenue on the mainframe remains undeniable.
- 36% say more than half of business revenue runs on mainframe applications
- 21% say more than 75% of business revenue depends on IBM Z
What runs on the mainframe isn’t fringe workload, either. It’s batch processing, event-driven execution, and high-volume transactions — the systems that move global money and enforce data integrity. The mainframe isn’t nostalgic infrastructure. It’s economic gravity.
2) Modernization is Disciplined, Not Reactive.
Mainframe environments aren’t frozen, nor are they chaotically modernizing. Changes to release editions are disciplined and sequenced.
- 34% are on the current z/OS release
- 40% run the previous release
- Only 3.2% plan decommissioning over the next three years
Most organizations plan to maintain or modernize in place rather than migrate away. Upgrades follow operational readiness and risk tolerance—not release-versioning hype. This is modernization through good governance.
3) Security Is Embedded and Incidents are Stable — For Now
In a year defined by cybersecurity headlines, the mainframe data tells a quieter, more positive story.
- 31% report a moderate decrease in incidents
- 30% report no change
That’s impressive considering that only 20% use MFA, 22% use Privileged Access Management, and 23% use SIEM/security monitoring integration. While raising questions of complacency, the data suggest that mainframe environments remain controlled-risk domains, even as hybrid complexity increases.
4) Hybrid Is the Operating Model, and Observability Is the Pressure Point.
Hybrid isn’t a transition phase. It’s the existing state, though it’s not perfected.
- 30% operate in a highly integrated hybrid model
- 39% report limited integration
- 43% cite observability/performance monitoring as the top hybrid challenge
Nearly 70% are actively integrating across platforms. This means the friction point isn’t “move off Z.” Rather, it’s visibility, coordination, and governance across distributed environments. As the mainframer already knows, complexity lives at the integration boundary, not at the system of record.
5) AI Expands Integration. It Doesn’t Replace Z.
Wary of overpromising and underdelivering, AI expectations are pragmatic rather than disruptive.
- 49% expect minor impact over the next 3–5 years
- Only 8% expect major change
- Top use cases: 29% anomaly detection and 26% security monitoring
Execution may shift toward public cloud or hybrid hosting, or remain as is. But organizations believe AI workloads will still depend on mainframe data. Admittedly, AI increases integration pressure, but it doesn’t eliminate the platform that holds the data.
0 Comments