Think 2026: IBM Positions Mainframe Environments as Governed AI Infrastructure
At its Think 2026 conference, IBM unveiled a broad expansion of enterprise AI, automation, and hybrid cloud capabilities, with several announcements aimed directly at organizations running mission-critical workloads on IBM Z and LinuxONE systems.
The company positioned the announcements around what CEO Arvind Krishna called a new “AI operating model” for enterprises, focused on governance, automation, real-time data access, and operational resilience across hybrid infrastructure.
For mainframe customers, one of the most notable announcements was IBM Z Database Assistant, now available in private preview. The AI-powered workspace is designed to help Db2 and IMS database administrators monitor performance, automate routine tasks, and optimize configurations across complex IBM Z environments.
The announcement aligns with growing industry concerns around the mainframe skills gap and the increasing operational complexity of large enterprise environments. IBM said the tool is intended to help administrators manage systems more efficiently while reducing manual workload.
IBM also announced IBM zSecure Secret Manager, planned for availability in June 2026. The product adds enhanced security automation for RACF environments and integrates with IBM Vault Self-Managed for IBM Z and LinuxONE to streamline certificate lifecycle management.
Security and governance were recurring themes throughout the conference. IBM emphasized that enterprises deploying AI systems increasingly require the same levels of auditability, policy enforcement, and operational control traditionally associated with core transaction infrastructure.
The company also introduced the next generation of IBM watsonx Orchestrate, described as an “agentic control plane” for managing large numbers of AI agents across enterprise environments. IBM said the platform is designed to provide centralized governance, policy enforcement, and accountability as organizations scale AI deployments.
Several additional announcements focused on data integration and hybrid infrastructure, areas closely tied to enterprise mainframe environments.
IBM said its recent acquisition of Confluent will strengthen real-time data streaming capabilities across hybrid environments through Kafka- and Flink-based technologies integrated with watsonx.data. The company also announced new context and governance capabilities for watsonx.data intended to help enterprise AI systems reason over operational data while maintaining explainability and runtime governance controls.
Hybrid operations and infrastructure automation were another focus. IBM unveiled the IBM Concert platform, an AI-powered operations platform designed to correlate signals across applications, infrastructure, networks, and security tools into a unified operational view.
IBM also introduced IBM Sovereign Core, now generally available, aimed at organizations operating in regulated industries or across multiple jurisdictions. Built on Red Hat OpenShift and Red Hat AI technologies, the platform is designed to embed governance and compliance policies directly into infrastructure operations while supporting workload portability across hybrid environments.
Taken together, the announcements reinforce IBM’s broader strategy of positioning the mainframe not as an isolated platform, but as a governed, secure foundation for enterprise AI and hybrid operations.
Source: IBM
Rocket Software Completes Acquisition of Vertica®, Expanding High-Performance Analytics and AI for Mission-Critical Systems
Rocket Software has announced the completion of its acquisition of Vertica®, an enterprise-grade analytics database platform, from OpenText. The acquisition advances Rocket’s strategy to bring intelligence to modernization by combining trusted core systems with high-performance analytics and AI to help enterprises unlock the full potential of their data. The acquisition brings more than 600 global customers and 170 employees to Rocket’s global organization.
“For more than 35 years, Rocket Software has been the modernization partner of choice for global Fortune 500 companies,” said Milan Shetti, president and CEO of Rocket Software. “With Vertica, Rocket is advancing the next phase of modernization by enabling customers to unlock the power of their enterprise data.”
Strengthening the modernization journey with Vertica
The acquisition of Vertica enhances Rocket Software’s ability to support customers along their modernization journey, including:
- Running Advanced Analytics and AI on Trusted Systems. Vertica adds high-performance analytics and AI capabilities to Rocket Software’s modernization portfolio, complementing core systems modernization with advanced data insights. This allows enterprises to run next-generation analytics and GenAI directly on trusted data sources, accelerating innovation and adoption.
- Supporting Broad Deployment Needs. Vertica’s flexibility across cloud, on-premises, and hybrid environments allows Rocket Software to serve customers wherever they are on their modernization journey.
- Strengthening Rocket Software’s Data Portfolio. By combining Vertica’s advanced analytics and data warehousing with Rocket® DataEdge™ and Rocket® ContentEdge™ solutions, which connect, govern, and activate enterprise data across complex hybrid environments, Rocket Software is creating an integrated data platform that expands insights, accelerates decision‑making, and strengthens the value of its data modernization portfolio.
- Accelerating Future Integrations. Building on Rocket Software’s 2024 integration of OpenText’s Application Modernization & Connectivity (AMC) business, this acquisition is expected to accelerate the integration of Vertica’s technology, enabling faster delivery of value to customers across Rocket Software’s global portfolio.
- Delivering Exceptional Customer Outcomes. With a 93.7% customer satisfaction rating, Rocket Software combines tailored support, deep industry expertise, and a consultative approach to help enterprises achieve success as they modernize and adopt AI-driven solutions.
Source: Rocket Software
Kyndryl unveils agentic AI capability that proactively prevents IT outages and accelerates recovery for enterprise customers
Kyndryl has unveiled a new patented capability in Kyndryl Bridge, the company’s AI-powered, open integration platform, that is enabling customers to automatically detect and resolve IT risks before they escalate into business-impacting outages.
Kyndryl’s prediction and prevention capability has been deployed on Kyndryl Bridge and is providing AI agent-assisted support to the more than 1,400 customers using Kyndryl Bridge. Kyndryl Bridge generates more than 16 million AI insights each month, has demonstrated a reduction in IT incidents by up to 50%, and drives an aggregate $3 billion in annual customer savings from avoided impact events and planned maintenance costs.
“By embedding AI agents in Kyndryl Bridge for proactive risk detection, we are transforming IT operations from reactive outage recovery to proactive, evidence based prevention,” said Xerxes Cooper, Global Leader, Kyndryl Delivery. “Correlating millions of observability signals across applications and deep infrastructure helps our customers see and resolve issues before they ever feel them.”
This proactive approach is powered by AI-agent assisted root cause analysis within Kyndryl Bridge, enabled across more than 200,000 customer devices to identify the underlying conditions that commonly precede outages. By accelerating analysis that once required extensive manual investigation, the platform enables teams to surface actionable insights faster – supporting earlier intervention and reducing the impact of complex incidents across hybrid and multi-vendor environments.
At scale, this advanced capability radically reduces the time required to complete root-cause analysis of major IT incidents, allowing organizations to complete reports in hours instead of weeks. Kyndryl experts review and validate the generated insights for operational context and alignment with customer environments.
Predictive detection and prevention of failures
Kyndryl’s new prediction and prevention feature brings evidence-based intelligence to enable predictive failure detection within IT operations by extending unified observability across a customer’s full IT landscape.
The patented feature dynamically identifies patterns that matter, and validates causal relationships between application slowdowns, infrastructure contention, configuration changes, and operational events. It does so by analyzing and delivering insights into how small anomalies accumulate and propagate across IT layers. This transforms IT operations from reactive recovery to proactive prevention, enabling teams to intervene early and reduce downtime across complex, multi-vendor environments.
Customer impact and availability
Customer engagements show encouraging results with accelerated detection and improved accuracy of issue prevention that may have led to business downtime. This capability handles early detection at scale for 10 million-plus incidents annually, and has demonstrated upwards of a 90% reduction in mission-critical production outages for certain customers.
Source: Kyndryl
Precisely Advances Agentic-Ready Data with a New AI Agent, Data Product Marketplace, and MCP-Enabled APIs
Precisely has announced new capabilities in its Data Integrity Suite to help organizations build, share, and use Agentic-Ready Data, the highest quality data that is integrated, governed, and enriched for AI, automation, and analytics initiatives across the enterprise. The release introduces a Data Integration Agent that joins the Gio™ AI Assistant, a data product marketplace integrated into the Data Integrity Suite through a partnership with Huwise, and expanded APIs now accessible through a Precisely-hosted MCP (Model Context Protocol) server.
Most enterprises are eager to move from AI experimentation to full implementation, yet inconsistent, non-compliant, and siloed data continues to be a bottleneck to forward progress. The latest enhancements to the Precisely Data Integrity Suite address these challenges by helping organizations build reliable data pipelines, publish trusted data products, and make contextualized data directly usable by AI systems and automated workflows.
The latest enhancements to the Precisely Data Integrity Suite include:
- Data Integration Agent helps teams design and configure data replication pipelines by handling setup, schema mapping, and validation tasks – reducing manual effort, improving consistency, and accelerating time‑to‑value. This agent joins the previously announced Gio AI Assistant and a growing collection of specialized AI agents for data quality, enrichment, and more.
- Data Product Marketplace, available through a partnership with Huwise, the leading provider of data marketplace solutions, enables organizations to publish and share trusted data products for internal and external use, so teams can reuse high‑integrity data across the business, collaborate with partners, and power analytics and AI without rebuilding data for each project.
- New APIs for Data Integration, Data Quality, and Data Catalog provide programmatic control over data pipelines, quality rules, and metadata – enabling automation and integration into AI-driven workflows.
- The Precisely-hosted Model Context Protocol (MCP) Server extends the Data Integrity Suite APIs, enabling AI agents and tools to securely discover, access, and use these capabilities without custom integrations. This builds on Precisely’s previously released MCP server, which focused on location intelligence and data enrichment APIs.
Together, the capabilities in the latest Data Integrity Suite release give organizations a more direct, governed path from raw data to Agentic‑Ready Data. With expanded APIs, a Precisely‑hosted MCP Server, new agent capabilities, and a governed data product marketplace, teams can use trusted data across applications and AI workflows, without the custom engineering and manual prep that typically slow AI initiatives.
For example, one of the user’s data consumers can now ask a natural-language question like ‘What are the quality scores for our customer records?’ directly through an MCP-compatible client and receive the information without writing custom API calls or routing the request through a data engineering team.
“Organizations are eager to scale AI, but data readiness remains the biggest obstacle,” said Matt Waxman, Chief Product Officer at Precisely. “With the latest release of the Data Integrity Suite and our partnership with Huwise for the Data Product Marketplace, we are helping customers turn their data into a trusted, reusable asset that can directly power AI applications and agent‑driven workflows.”
Source: Precisely
The AI Patching Crisis Hits the Enterprise: What “Fragnesia” Means for Linux on IBM Z
An ominous high-level Linux kernel vulnerability named “Fragnesia” has been uncovered by security firm Zellic using AI-agentic auditing tools, marking the third major local root flaw discovered in just two weeks. This rapid-fire succession of bugs—following hard on the heels of “Copy Fail” and “Dirty Frag”—signals a paradigm shift that enterprise infrastructure teams cannot afford to ignore.
Generative AI engines like Claude Mythos and OpenAI Daybreak are discovering open-source security vulnerabilities far faster than human maintainers can write, test, and push downstream patches. For organizations running Linux on IBM Z (z/VM or Integrated Facility for Linux/IFL), this acceleration of weaponized flaws fundamentally threatens the traditional window of vulnerability management.
While Fragnesia is technically categorized as a local privilege escalation bug (with a severe CVSS score of 7.8), its architecture makes it a significant threat to high-density mainframe environments. The flaw abuses a logic bug in the network subsystem to corrupt the kernel page cache of read-only files, bypassing standard security isolation without needing complex timing tricks. In a multi-tenant cloud or enterprise mainframe environment where thousands of untrusted containers or virtual machines share a monolithic Linux kernel, a functional proof-of-concept exploit could enable an attacker to escape a container, gain full root on the host system, and compromise adjacent corporate workloads.
Key Takeaways for Mainframe Systems Engineers
- The Shared-Kernel Risk: Because Fragnesia allows reliable local privilege escalation without race conditions, it completely undermines the logical isolation of containerized microservices running on shared IFLs.
- AI Outpacing Upstream Patching: Mainframe environments thrive on rigid change management and meticulously validated patch cycles. However, AI-driven bug discovery is introducing zero-days at a velocity that threatens to break traditional enterprise maintenance windows.
- Exploitation Automation: A functional proof-of-concept exploit already exists that replaces standard system commands with a malicious stub, making the vulnerability highly predictable and easily weaponized once an attacker gains initial local access.
- Immediate Mitigation Strategy: While an upstream kernel fix has been engineered to tighten fragment handling, it has not yet been fully packaged by commercial enterprise distributions (such as RHEL or SLES). Systems administrators must track their specific vendor streams and prepare for immediate, out-of-band kernel updates to protect core business profiles.
Source: ZDNet
0 Comments