IBM focused product innovation news at its annual conference on how to enable an “AI operating model” that integrates data, agents, automation and hybrid cloud, delivering AI to the core of the business and beyond a collection of AI projects.
A private preview of a new iteration of Watsonx Orchestrate, a public preview of the concert operations platform, and general availability of the IBM Bob agent developer product and Sovereign Core software platform are some of the biggest news announced at IBM Think 2026.
The Armonk, New York-based technology vendor shared product innovation news at its annual conference with a focus on how to enable an “AI operating model” that integrates data, agents, automation, and a hybrid cloud to bring AI to the core of the business and move beyond a collection of AI projects to enable connected AI systems. Thoughts continue until Wednesday in Boston.
IBM’s AI opportunities extend to service partners, reseller partners, build partners, ISVs and startups, IBM Chairman, President and CEO Arvind Krishna said in response to questions from CRN during a virtual press conference on Monday.
Significant revenue opportunities exist for system integrators and service partners in connecting data sources to unlock AI agents and AI operating models.
“That’s the value that businesses need, and it’s not just a chatbot or a consumer-facing application,” Krishna says. “Those are important, but they only provide the first 20% of the value. The next value comes from implementing hybrid cloud technology.”
[RELATED: IBM Q1 2026 Earnings: CEO Krishna Says AI Is A Tailwind To The Business]
IBM announces AI operating model
Krishna encouraged solution providers to look to the OpenShift hybrid cloud platform, IBM subsidiary Red Hat, and even the IBM Sovereign Core announced at Think 2026. He also recommended investing in increased automation to eliminate human labor complexity wherever possible and move value to real-time.
Rob Thomas, IBM’s senior vice president and chief commercial officer of software, told CRN during a virtual press conference that the vendor is exploring multifaceted partnerships, and has put work with tax and consulting giant EY on hold, for example.
EY can use IBM Watsonx Orchestrate and IBM Bob to power new tax products, bring services and advice to market, and bridge that product to other products through its technology platform.
“This is an example of a service partner taking the next step in AI adoption by implementing AI as one of their solutions,” Thomas said. “And together we built a collaborative work commitment with them.”
Bo Gebbie, president of Hamel, Minn.-based IBM solutions provider Evolving Solutions, 162 on CRN’s 2025 Solution Provider 500, told CRN in an interview that IBM’s recent acquisitions HashiCorp, IBM Bob, and automation tools have proven popular among customers recently.
“You think about increased labor costs. You think about AI workloads. The automation side impacts all of that,” Gebbie said.
IBM Fusion infrastructure for AI and containers brings enterprise-grade performance across hybrid environments, and integrating compute, storage and networking is also a huge opportunity for evolving solutions, he said. “We are making a significant push for skills and investment in that area.”
New IBM Agentic AI, Watsonx, Concert release
Thomas said in a press conference that the new version of Orchestrate, which is in private preview, will be home to more third-party agents from companies such as ServiceNow, Salesforce, and Adobe.
“Orchestrate is no longer just about IBM technology,” says Thomas. “This is the best agent technology of any company in the world.”
IBM said the Bob Premium Package for Z, which extends Bob functionality to mainframe environments, is also in private preview.
IBM is moving the context of Watsonx.data feature into private preview, giving users an open, federated context layer for reliable AI inference on business data. Users can apply semantic meaning, apply governance at runtime, and provide explanations for decisions. Watsonx.data gets commonly available integrations with Confluent Tableflow and Flink for real-time event streaming. Private preview of IBM Z Database Assistant provides Db2 and IMS database administrators with an AI-powered workspace to monitor performance, automate tasks, and optimize configurations.
A separate private preview brings Watsonx.data GPU-accelerated Presto to some users, reducing certain workload costs and processing times for large datasets.
Among the public previews unveiled during Think 2026 is the IBM Concert AI-powered operations platform. Concert provides users with a single view of signals across applications, networks, infrastructure, and costs without replacing existing tools.
The concert promises cross-domain understanding to prevent silos and bring critical data to the surface, Thomas said, and its importance grows as the world worries about new risks posed by the myth of Claude’s creator Antropics. “Every customer is looking for how to solve a vulnerability problem that I’m not aware of,” he said. “The only answer to that in today’s market is concerts.”
Concert Secure Coder, for security management built into developer workflows, is also in public preview. Secure Coder is available for IBM Bob and VS Code and generates automatic repairs for vulnerable code and patches middleware, packages, images, and more.
HashiCorp Platform (HCP) Terraform, powered by Infragraph, is in public preview and enables unified infrastructure visibility through a centralized, event-driven knowledge graph. IBM says users will connect data across cloud environments, infrastructure-as-code workflows, security tools, and operational platforms.
IBM has made Bob agent developer generally available. Bob enables you to build AI agents with built-in security and cost.
IBM also made Vault 2.0 generally available with AI-powered analysis for rapid triage of leaked secrets, dynamic short-lived credentials across major cloud providers, and automatic secret rotation.
The vendor plans to make Secure Secret Manager available in June to enhance mainframe environments with resource access control capabilities. IBM says the manager is integrated with Vault Self-Managed for Z and LinuxOne.
IBM begins general availability of Sovereign Core
IBM is making Sovereign Core generally available, providing users with an AI-enabled sovereign environment and a software platform for controlling infrastructure, operations, and systems. The platform, which has been in preview for several months, offers a way to deploy technology in a border-isolated environment or on-premises and only be managed locally, Thomas said.
According to the vendor, Sovereign Core aims to help AI users create consistent and auditable AI results, allowing cutting-edge technologies to be deployed into operational reality faster. Sovereign Core aims to enable operational sovereignty and sovereignty over data at rest, in use, and in motion. We also focused on technology sovereignty, avoiding vendor lock-in through an open, modular architecture. Users then gain sovereignty over the AI, controlling where their models run and how they manage inference.
The platform control plane extends full control over configuration, operations, and lifecycle management. Inbound identity, encryption, and data services give customers control over access, secrets, keys, logs, and audit evidence. Sovereign Core has continuous compliance monitoring and evidence generation and is ready for real-time audits. It also comes preloaded with regulatory frameworks to quickly define compliance posture across regions and industries.
Sovereign Core enables management of data, models, inference, and agent behavior, control of where AI processing occurs, and governance of access and updates, all of which are important for highly regulated sectors such as governments, the public sector, regional cloud providers, and enterprises, IBM said.
The platform already prioritizes third-party and open source software and services from vendors such as AMD, Palo Alto Networks, Cloudera, Dell, Intel, MongoDB, Mistral, and Elastic, as well as channel partners such as HCL and Atos.
According to IBM, users provision CPUs, GPUs, and AI inference environments using standardized templates and automated configuration profiles for consistent workload deployment and management.
