For decades, the mainframe has been defined by the transactions and data enterprises trust it to process. AI is extending that role.

IBM Institute for Business Value research found that 75% of executives expect mainframe-based applications to remain central to digital transformation, while 60% say mainframe platforms are essential to enabling AI innovation.

The mainframe increasingly supports intelligence within the operational flow, combining enterprise data, embedded business logic and AI while preserving the resilience, security and scalability organizations already depend on.

Bringing intelligence into the transaction

Some of the strongest opportunities for AI sit alongside the data and business logic already operating on the mainframe.

Decades of transactional history provide context for fraud detection, risk assessment, operational optimization, personalized experiences and next-best-action recommendations. Bringing intelligence closer to this data reduces latency and limits unnecessary movement of sensitive information across platforms.

Specialized processing technologies such as IBM Telum II and Spyre accelerators support AI workloads closer to core transactions without relying on conventional processing capacity in the same way.

AI is also becoming more deeply embedded in mainframe system software. Performance and capacity management, event correlation and workload management increasingly draw on AI-driven analysis.

AIOps extends that visibility further. Leaders can use operational intelligence to understand areas such as system performance, capacity, total cost and internal chargeback. Intelligence can then begin to shape both business transactions and the operation of the platform itself.

Modernizing around business capabilities

AI also changes how enterprises can approach modernization.

GenAI accelerates the analysis of legacy applications, helping uncover business rules and document dependencies that previously required significant manual effort to understand. Teams can build a clearer picture of what an application does before deciding how it should evolve.

Our work on AI-led mainframe modernization takes this business-led approach.

Leaders should focus modernization decisions on the capabilities that create the greatest value. Others may contain business logic that remains highly valuable and should be preserved while the surrounding architecture evolves.

This selective approach accelerates innovation while reducing unnecessary disruption to mission-critical processes.

Introducing intelligence without weakening resilience

AI embedded in core processes needs the same operational discipline applied to other mission-critical systems.

Organizations should begin with clearly defined use cases in which AI supports decisions, then expand automation as evidence and confidence grow. Observability, security controls, auditability, rollback mechanisms and policy-based guardrails should be built into the implementation.

Early automation should focus on read-only activities such as display commands and read-only batch operations. Changes to production systems should remain attributable to an identified user or role, while high-impact actions continue to require human authorization.

Keeping AI close to trusted enterprise data may also help where latency, privacy or compliance requirements make data movement undesirable.

These controls allow organizations to increase decision speed and automation while preserving the reliability and security associated with the mainframe.

Measuring an intelligent core

Technology metrics provide only part of the picture.

Leaders should also assess decision speed, customer experience, operational risk, resilience and the time required to launch new products or services. Those measures show whether intelligence embedded in the core is changing business performance.

Practical examples include using Agentic AI with Ansible playbooks to resolve established problems or applying AIOps to identify suboptimal configurations before they affect performance.

An intelligent enterprise core connects trusted data, business logic and AI directly to operational decisions. Its value becomes visible when the organization can act faster, respond to change with greater confidence and create better outcomes from the systems already at the center of the business.

Modernization then becomes a continuous process of increasing the intelligence and business value of the core.