Red Hat pushes agentic AI into production with expanded platform and partnerships

Enterprise organizations are racing to deploy artificial intelligence faster than their budgets and infrastructure can support. Red Hat Inc. is tackling this gap by helping customers move AI from experimental phases into production systems that generate real business value.

The company presented its strategy at Red Hat Summit 2026 in May, emphasizing expanded platform features, operational automation and an open-source model built on community participation. "I think everybody here loves that because that does lead to sovereignty as well," said Rob Strechay, principal analyst at theCUBE, during the event. "They're actually building on top of open source and enhancing it in a way that provides more value in supporting it, which is key to their success."

New platform and services for hybrid AI workloads

Managing AI systems across hybrid cloud setups creates operational complexity that many enterprises struggle to handle. Red Hat introduced Red Hat AI 3.4, an updated platform designed to support agentic AI and large-scale inference operations running across hybrid environments.

The company also unveiled two new managed services. Red Hat AI Inference brings real-time AI inference capabilities directly into customer production systems, while Red Hat OpenShift Virtualization Service on IBM Cloud enables organizations to run virtual machines at scale. Both services aim to accelerate the shift toward hybrid cloud infrastructure.

"What's really going to drive inference demand exponentially is AI agents," said Joe Fernandes, vice president and general manager of Red Hat AI. "We provide a platform where customers can deploy and manage their AI agents across a hybrid infrastructure environment."

Red Hat also released Ansible Automation Platform 2.7, which connects AI-generated insights to controlled infrastructure actions. The platform allows authorized staff at branch offices and factory floors to manage servers and networks within preset governance boundaries.

"Customers are now able to use their own data to calculate savings and efficiencies," said Sathish Balakrishnan, vice president and general manager of Red Hat's Ansible business unit. "Every action, whether it's triggered by humans, an event or an AI agent, passes through the same [access controls], approval gates, audit trails, content signing and credential management."

Addressing the AI budget and ROI challenge

Spending on AI cloud services is climbing sharply. According to theCUBE Research, analysts project AI cloud spending could reach $2.5 trillion by 2026, expanding at roughly 44% annually. Yet many organizations struggle to convert that investment into tangible results. Research indicates that while 50.7% of organizations use public AI tools like ChatGPT and Copilot, only 20.2% have deployed enterprise-wide AI systems built on a governed framework.

"Most people are token consumers and that is part of the reason they're trying to figure out how do they get to production," Strechay said. "How do they get out of these death-by-pilot situations and become token producers so that they can offset the cost or at least lower the cost so they can get to the [return on AI investment] of those applications? You can use [OpenShift] infrastructure along with things like vLLM and some of the other things in Red Hat AI to get and become efficient in how you produce your tokens."

Red Hat partnered with Advanced Micro Devices, Inc. to help enterprises rethink AI economics. Rather than purchasing the latest high-end hardware, organizations may achieve better results by matching compute resources to specific workload requirements, according to John Hampton, corporate vice president of global enterprise technical sales at AMD.

"So many of these enterprises just ran out and bought big [graphics processing unit] clusters because they knew they had to solve for AI," he said. "Enterprises are coming to us and saying, 'This is exploding. It's good that we're using AI, but we can't afford this anymore.'"

Red Hat also joined with Intel Corp. to guide customers toward hardware and software combinations suited to their specific needs. As organizations seek to lower inference expenses, both companies are promoting the value of balancing CPUs and GPUs rather than relying exclusively on one approach.

"As we've gone through this with our customers in the industry, we've seen that people often have just assumed, 'I've got a hammer. I need the nail to hit it with,'" said Bill Pearson, vice president of data center and AI at Intel. "Once they take a step back to say, 'Wait a minute. I have these [central processing units] in my data center' or 'I need to figure out how to balance the right number of CPUs with the right number of GPUs to achieve that outcome I'm looking for,' they're actually going to get better results at a better price point for delivering lower-cost tokens."

Strategic partnerships span enterprise to edge computing

Red Hat announced multiple collaborations at the summit. Red Hat AI Factory, developed with Nvidia Corp., includes support for Nvidia's Blackwell architecture and the forthcoming Vera Rubin platform, plus participation in Nvidia's OpenShell project for AI agent sandboxing and secure execution.

"Moving AI from corporate experimentation to an industrial engine requires a sovereign, consistent foundation across the hybrid cloud," said Chris Wright, chief technology officer and senior vice president for global engineering at Red Hat. "Through our strategic co-engineering efforts with Nvidia, Red Hat provides the architectural control and open source innovation enterprises need to scale agentic AI with confidence."

Additional partnerships target agentic AI and edge infrastructure. Data activation company Boomi LP is working with Red Hat on an integrated stack for deploying agentic AI at scale. Panasonic Connect and Voyager are collaborating on edge computing solutions and space edge micro data centers, respectively.

Red Hat's expanding partner network reflects the broadening range of business use cases for agentic AI. The real challenge lies in determining how these capabilities apply to individual organizational needs, according to Matt Hicks, president and chief executive officer of Red Hat.

"From the board level down … you have a feeling of, 'We have to be able to harness what this technology can do,'" said Matt Hicks. "The people that will be the most successful with AI are the people that have the most questions. People that know business, they know their domain. Those are the ones [that] have no barriers to what they can do and change. It's just about getting started."

Sovereign AI and governance capabilities expand

Sovereign AI—giving organizations greater control over their data and infrastructure—has become increasingly critical for enterprises. Gartner Inc. forecasts that the sovereign cloud infrastructure-as-a-service market will grow at a compound annual rate of 36% through 2028.

Red Hat is broadening sovereign cloud support while allowing customers to retain existing infrastructure investments. The company is collaborating with Core42, which focuses on sovereign-by-design infrastructure, and Sopra Steria, a longstanding European partner on sovereign AI work.

Red Hat is also strengthening software security as enterprises adopt agentic AI. Fedora Hummingbird Linux is a container-native operating system built for agent-first development with an automated software delivery pipeline. Red Hat Hardened Images offers a collection of minimal container images designed to reduce security vulnerabilities and protect software supply chains.

Sovereign AI and governance mechanisms are now essential for giving customers control and choice, according to Ashesh Badani, senior vice president and chief product officer of Red Hat.

https://www.youtube.com/embed/5xzfzuqK3Ow?feature=oembed

"Whether an organization is meeting jurisdictional mandates or reclaiming its data from proprietary silos, we are providing the capabilities and platforms to build a more self-determined future," he said.