Can Kyndryl’s agentic AI approach finally modernize the IBM Z mainframe?

Kyndryl (NYSE: KD), one of the largest global providers of enterprise infrastructure services, has launched a new suite of agentic AI-powered capabilities for IBM Z systems, targeting the evolving needs of large enterprises looking to modernize their core mainframe environments. Announced on November 24, 2025, the initiative leverages agentic AI orchestration, hybrid IT integration, and the open Kyndryl Bridge platform to streamline operations, automate decision-making, and enable faster development across application and infrastructure lifecycles.

The launch is designed to help customers running IBM z/OS platforms deploy more intelligent, autonomous workflows using agentic AI principles—systems that sense, reason, and act autonomously across IT environments. Kyndryl is positioning this capability as a foundational step in turning traditional mainframes into active engines of AI-enabled business innovation.

This new offering integrates the company’s existing services portfolio with advanced AI features from IBM watsonx Assistant for Z and other tools in the AI ecosystem. It includes the introduction of the Kyndryl AI Assistant for Z, a knowledge-based platform trained on decades of Kyndryl’s mainframe operational experience, which has now been embedded within Kyndryl Bridge to enhance customer support, platform management, and skills development.

How Kyndryl is using agentic AI to transform legacy mainframe architectures into active AI participants

At the center of the offering is Kyndryl’s Agentic AI Framework, a model designed to help orchestrate and manage autonomous agents across complex enterprise workloads. These agents operate within the boundaries of client mainframe systems, continuously monitoring processes, identifying inefficiencies, and executing proactive interventions without the need for manual triggers.

The agentic AI services are tightly integrated with the Kyndryl Bridge platform, enabling clients to gain predictive insights, real-time observability, and intelligent automation across their hybrid environments. As enterprises increasingly adopt AI and digital transformation strategies, Kyndryl’s new capabilities offer a bridge between legacy infrastructure and next-generation computing paradigms.

This architectural alignment also allows for closer proximity between AI inferencing workloads and mainframe data. By enabling inferencing closer to transactional data sources, enterprises can reduce latency, improve fraud detection speeds, and better support use cases such as insurance claim processing or customer transaction analytics. This proximity has the potential to drastically enhance performance while maintaining the reliability and security advantages that mainframes are known for.

Why mainframe modernization needs AI now, and how Kyndryl plans to meet that demand

According to findings from Kyndryl’s 2025 State of Mainframe Modernization Survey, 88 percent of enterprise respondents have either already implemented or plan to implement AI capabilities, including agentic and generative AI, within their mainframe environments. Among the top motivations cited were the need to improve business agility, reduce human error through repeatable automated processes, and generate significant cost savings.

The report projects that AI adoption in mainframe environments could drive $12.7 billion in cost reductions and create $19.5 billion in incremental revenues over the next three years across surveyed enterprises. However, the study also revealed a persistent challenge facing these organizations: 70 percent of respondents reported difficulties in finding talent with the right mix of AI and mainframe expertise to lead these transformations.

Kyndryl’s response to this growing talent gap is its AI Assistant for Z. Trained on decades of institutional experience and deployed as a core component within Kyndryl Bridge, this assistant is designed to fill knowledge and expertise voids by providing contextual recommendations, lifecycle management insights, and AI-driven system alerts to support operational continuity. These tools are expected to significantly reduce the time needed to detect and resolve issues, speed up development cycles, and allow internal teams to focus on higher-value innovation activities.

Hassan Zamat, Global Practice Leader for Core Enterprise and zCloud at Kyndryl, noted that agentic AI represents a turning point in how mainframes are managed and perceived. He stated that by pairing delivery practitioners with AI orchestration, enterprises can build trust in the mainframe as not just a legacy cost center, but as a strategic platform for growth, risk mitigation, and innovation.

How Kyndryl’s AI platform connects to IBM’s watsonx and wider automation ecosystem

The rollout is also strategically aligned with IBM’s broader AI infrastructure strategy. IBM watsonx Assistant for Z has been integrated as a core component of Kyndryl’s offering, providing clients with additional tooling for intelligent orchestration and lifecycle management of their mainframe environments.

Skyla Loomis, General Manager of IBM Z Software, highlighted that the collaboration with Kyndryl aims to empower enterprises to bring advanced AI capabilities to the systems that power their critical operations. She emphasized that AI integration with mainframe environments can significantly improve decision-making, boost operational agility, and enable predictive infrastructure planning.

By building on IBM’s tools and extending its own Bridge platform, Kyndryl is providing customers with an interoperable AI layer that sits across multiple systems. This is particularly relevant for hybrid IT strategies, where enterprises are balancing workloads across on-premise mainframes, private clouds, and hyperscaler ecosystems.

The new services are not limited to enhancing performance alone. They are also designed to facilitate compliance, reduce risks, and improve incident response. Through prebuilt orchestration templates and real-time monitoring capabilities, enterprises can automate issue remediation, proactively adjust resource allocation, and align IT performance with changing business priorities.

Why agentic AI is emerging as a crucial lever in modern enterprise infrastructure

The rise of agentic AI is closely tied to the growing complexity of enterprise IT systems. With thousands of interdependent processes running across hybrid cloud and mainframe environments, manual intervention is increasingly impractical. Agentic AI offers a way to abstract this complexity, enabling systems to operate more autonomously, respond faster to anomalies, and reduce the burden on human operators.

In the context of mainframes, this is particularly important. These systems are still heavily relied upon for mission-critical operations in banking, insurance, healthcare, and government sectors, but are often hampered by aging skills bases and siloed application architectures.

Kyndryl’s agentic AI services effectively give these systems a modern upgrade, allowing them to interact dynamically with other platforms, respond to evolving security threats, and deliver intelligent, explainable automation. With many enterprise clients facing the dual challenge of maintaining mainframe reliability while enabling innovation, this model could offer a sustainable pathway forward.

The inclusion of real-time fraud detection and AI-enhanced customer interaction monitoring also positions Kyndryl’s solution as relevant to digital-native enterprises seeking better visibility across user behavior, transaction anomalies, and real-time decisioning.

What the future holds for mainframe-based AI services and Kyndryl’s enterprise roadmap

Kyndryl’s investment in mainframe modernization, particularly through AI-powered delivery, reflects a broader trend in the IT services industry: the repositioning of legacy infrastructure as a value-generating asset rather than a liability. As hybrid environments become more normalized, having unified orchestration across mainframes and cloud systems is emerging as a competitive differentiator.

For Kyndryl, this launch also suggests a maturation in its post-spinout strategy. Since its separation from International Business Machines Corporation in 2021, Kyndryl has focused on expanding its capabilities beyond traditional managed services. With its Bridge platform, cloud ecosystem partnerships, and now a robust agentic AI service offering, the enterprise IT services firm is building a modular, AI-optimized approach to infrastructure management that could serve as a blueprint for large organizations facing similar modernization challenges.

Analysts watching the infrastructure-as-a-service space believe that Kyndryl’s push into intelligent automation, particularly within core environments like IBM Z, could help it defend and grow its market share amid increased competition from hyperscalers, consulting firms, and other system integrators pivoting toward AI operations.

As enterprises continue to prioritize risk mitigation, operational agility, and IT resilience, the demand for platforms that deliver explainable, actionable, and autonomous insights is only expected to rise. Kyndryl’s agentic AI services represent one of the more comprehensive entries into this evolving segment, especially for organizations with significant mainframe dependencies.

What are the key takeaways from Kyndryl’s agentic AI launch for IBM Z mainframes?

• Kyndryl has launched a new suite of AI-powered services tailored for IBM Z, combining agentic AI, hybrid IT orchestration, and mainframe expertise.

• The offering integrates with Kyndryl Bridge and IBM watsonx Assistant for Z to enable predictive automation and intelligent lifecycle management.

• A core feature is the Kyndryl AI Assistant for Z, a knowledge-based platform trained on decades of mainframe operational experience.

• The agentic AI framework allows autonomous agents to manage workloads, detect anomalies, and execute interventions with minimal human input.

• Real-time decision-making is a major benefit, enabling applications such as fraud detection, insurance claim processing, and customer analytics.

• Kyndryl’s 2025 survey shows 88 percent of enterprises are using or plan to use AI on mainframes, aiming for $32.2 billion in value over three years.

• Talent shortages remain a challenge, with 70 percent of organizations citing difficulty in hiring multi-skilled professionals.

• The solution positions legacy mainframes as active AI participants, transforming them into innovation platforms rather than static infrastructure.

• Analysts view this as a strategic step in Kyndryl’s post-IBM evolution, aligning its portfolio with enterprise needs for resilient, automated IT.

• The launch comes as enterprises seek unified platforms for hybrid IT management that combine observability, compliance, and intelligent automation.