AI & Data Exchange 2026: Red Hat’s Michael Hardee on building a roadmap to scale adoption

Federal agencies have identified more than 3,000 potential use cases for artificial intelligence, but significant hurdles remain as they try to scale pilot projects into enterprisewide tools.

Michael Hardee, chief architect for law enforcement and justice at Red Hat, said agencies are stepping up their use of AI to meet their missions. But they’re still contending with longstanding technical and operational barriers, including legacy IT systems.

Despite these pressures, Hardee said he has seen strong leadership and technical expertise across federal agencies.

“There are some brilliant technical talent and tons of technical leadership there. They’re doing a really fantastic job of adopting this technology and bringing it to their mission sets,” he said during Federal News Network’s AI & Data Exchange 2026.

Expanding AI to the edge

Agencies are deploying AI capabilities outside traditional data centers, including in disconnected or resource-constrained environments. Hardee said that flexibility is especially important for personnel in the field.

“Whether it’s the warfighter, whether it’s our Border Patrol or Coast Guard out at sea, we want to make sure that they have the capabilities required to serve the mission,” he said.

Given the rapid pace of change, Hardee advised agencies to take a collaborative and forward-looking approach to AI adoption.

“Work with your partner ecosystem, number one. Don’t reinvent the wheel. There is so much happening and so much work is being done. A lot of answers to your questions are probably already there,” he said.

Hardee also stressed the importance of long-term planning.

“Plan for three years, plan for five years, plan for 10 years. Have a roadmap. Get a roadmap from your vendor ecosystem and also hold them accountable,” he said.

Navigating supply chain concerns

Agencies must grapple with external constraints, particularly supply chain issues affecting critical hardware needed to run AI workloads.

“Everyone hears about GPU and memory, and the issues that we’re facing as far as getting those to our customers. That challenge is very real. Sometimes it 12 to 24 months to actually get physical hardware to be able to host these workloads,” Hardee said.

Meanwhile, the pace of innovation in the AI market is creating hesitation among agencies trying to make long-term investment decisions.

Leaning into open source for transparency in AI

As agencies adopt AI more rapidly, they are also under pressure to ensure these systems are transparent, explainable and secure. Hardee pointed to open source approaches as an enabler of trust.

“Observability is the word of the day. I want to be able to see every part of your software,” he said. “Over the last couple years, there’s been a lot of conversation around software bills of materials. When an application is shipped, we want to understand the bill of materials, what is actually inside of this. And that translates to AI models as well.”

Open source tools and models give agencies the ability to inspect and evaluate AI systems before deployment.

“We’re providing tooling in the open source community as well as open source models that allow you to introspect and investigate and understand what’s in these models. The power of the open source community is just that,” Hardee said.

“We’re providing you the visibility to see what’s in there, see if there’s any biases. Understand what you’re getting into. Working out in the open is how we’re going to get this thing adopted. Because there can be no black boxes here. We can’t have bias, especially when we’re talking about providing information. We want to understand the information that we’re getting.”

Slowing down to speed up

As agencies expand AI use, that shift could allow federal employees to focus less on routine tasks and more on decision-making.

“We’re going to be able to bring data together to get that information to humans in the loop, to be able to empower them to make decisions, to have that decision advantage much faster.”

As agencies continue their AI journeys, Hardee said the combination of modernization, transparency and strategic planning will be critical to turning early successes into sustained mission impact.

“At its core, what we’re talking about is automation. The challenge we’ve really had with automation — even though there’s been tremendous adoption of automation across public sector — is that you have to take the time to build it,” he said. “Even though there’s loads of documentation and tons of developers that are able to do things quickly, it still takes time. It’s an investment.”

AI is changing that equation by speeding up development and enabling faster insights.

“AI is going to change the trajectory to where we are going to be able to automate things much faster.”

Discover more articles and videos now on the AI & Data Exchange event page.

Copyright © 2026 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.