Digital accessibility has long been treated as a compliance checkbox. But for government agencies, it’s rapidly becoming something else entirely: a test of public trust.

What’s at stake is clear: Too many agencies continue to see compliance as a single audit instead of an ongoing responsibility. Without future-conscious solutions the problem is twofold: Millions lose access to resources they are entitled to, and organizations are forced into remediation contracts that balloon into seven-figure commitments. The time to act is running out, according to Robin Braun (pictured, middle), vice president of AI business development, hybrid cloud, at Hewlett Packard Enterprise Co.

“If you’re not on top of it now, you’re not going to be prepared,” she said, referring to looming enforcement deadlines that will soon pressure every municipality over 50,000 residents to meet stricter accessibility mandates.

Braun and Jack Hogan (right), vice president of advanced growth technologies at SHI International Corp., spoke with theCUBE’s Rob Strechay for an exclusive conversation on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the rising urgency surrounding digital accessibility compliance and how artificial intelligence is helping agencies accelerate execution without inflating headcount and budgets. (* Disclosure below.)

AI scales digital accessibility without scaling headcount

Under Section 508 of the Rehabilitation Act and the Web Content Accessibility Guidelines 2.0 standard, agencies are legally obligated to ensure that online services are accessible to all citizens, Strechay noted. But years of legacy PDFs, outdated layouts and inconsistent archives have made remediation both costly and difficult to sustain. To break that cycle, HPE and SHI have partnered with AI software provider Kamiwaza Corp. on a turnkey remediation platform powered by Nvidia Corp. graphics processing units. Its agent-based system scans entire sites, identifies inaccessible elements across HTML, imagery, video and documents, and automatically generates structured fixes — creating significant savings for government agencies, according to Hogan.

“We’ve had conversations with city governments and county governments that have three-year, multimillion-dollar budgets just to get to a place where they are compliant,” he said. “By being able to speed that up to weeks and really tens of thousands of dollars, it radically reduces the tax burden that is funding many of these projects and allows those funds to be redeployed, to accelerate other areas of ability for citizens to access information, to be able to have better outcomes leveraging advanced technology.”

While remediation can be handled through AI, decision-making remains human-led. Municipal staff receive prioritized recommendations across layout, structure and markup — approving or editing in place before any changes are published. That approach ensures digital accessibility improvements are fast-tracked without undermining editorial control or public accountability. And unlike traditional engagements that stop the moment a contract ends, remediation continues in the background, according to Braun. 

“Agents don’t get tired,” she said. “They don’t call in sick. They don’t take a break. As you continue to iterate on the website so that it stays in compliance … we continue to do those checks. It’s not a one and done, which is what is so difficult when you start to think about doing this manually or doing it consulting-wise.”

Digital accessibility as the new data layer

Aaccessibility work is not merely regulatory cleanup — it is data restructuring at scale, according to Hogan and Braun. Every inaccessible PDF or untagged image is not only unusable to screen readers, but also invisible to search engines, internal analytics tools and future citizen-facing assistants. When agencies transform inaccessible material into structured formats, they are not just fixing pages — they are rebuilding their content repositories into machine-readable knowledge systems, Hogan added.

“When you don’t start with great foundational data and you’re not compliant with some of the requirements and the regulations and the laws, it actually starts your AI journey off in really challenging ways,” he said. “Having compliance around 508 is really just the foundation.”

That view reframes digital accessibility from policy maintenance to technical modernization. What begins as remediation becomes the unlocking of institutional knowledge — exposing decades of static documents to search, translation and automated retrieval. But structuring data isn’t enough — agencies must also protect it while making it usable at scale, especially as sensitive records migrate into AI workflows, Braun explained.

“Their system understands data sovereignty [and] data gravity so that we’re not exposing anything that shouldn’t be exposed,” she said. “However, we’re ensuring that the things that people are accessing are accessible to all.”

Here’s the complete video interview, part of an exclusive interview from SiliconANGLE and theCUBE:

(* Disclosure: Hewlett Packard Enterprise Co. sponsored this segment of theCUBE. Neither HPE nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE