How AI Is Helping Governments Make PDFs Accessible Before the 2026 Deadline

AI PDFs: Helping Governments Meet Accessibility Deadlines

State and local governments serving populations of 50,000 or more face a pressing deadline: 10 months remain to comply with new accessibility standards. These rules ensure that people with disabilities can access government information and services online. While progress has been made, one major challenge remains—making thousands of PDF files accessible to assistive technologies like screen readers.

Take Georgia.gov as an example. Its main website alone hosts about 50,000 PDFs and other static files such as PowerPoint decks. Approximately one-third of these files require remediation to meet accessibility standards, according to Jennifer Thom, senior director of data science at Code for America.

Accessibility Progress and the PDF Challenge

Over the past decade, state and local governments have improved significantly in web accessibility. Outdated designs have been replaced with clearer typography and user-friendly layouts. Many states, like Georgia, rank among the top in following accessibility standards.

However, the remaining accessibility issues tend to be technical and specific rather than cultural. The biggest hurdle now is PDF files. These documents often lack the proper structure needed for screen readers and other assistive devices.

Thom explains, “The last-mile problem people are experiencing is dealing with PDFs.” Many PDF files require detailed review and remediation, a time-consuming process for government agencies facing tight deadlines.

How AI is Supporting Accessibility Efforts

Code for America developed an AI-powered tool to assist governments in managing their PDF files. Georgia and Salt Lake City are currently using this tool to review and classify static files ahead of the 2026 deadline. Governments with smaller populations have an extra year to comply.

The tool looks like a file manager interface, letting users filter and sort documents by type or date. A machine learning algorithm categorizes files—agreements, agendas, brochures, and more. Meanwhile, a generative AI summarizes the file contents and flags which documents may be exempt from remediation, such as archival-only files.

While the tool does not fully automate PDF remediation, it helps agencies understand the scope of work involved. For instance, fillable forms often need to be recreated from scratch, while many PDFs only require markup improvements for screen readers.

Legal Context and Practical Benefits

The accessibility standards come from the World Wide Web Consortium (W3C), but the driving force is the Americans with Disabilities Act (ADA). Non-compliance can be legally interpreted as discrimination, increasing the risk of lawsuits.

The ADA emphasizes serving all community members, including those with visual, auditory, physical, speech, cognitive, and neurological disabilities. Making online government services accessible also benefits agencies by improving efficiency—for example, accessible tax forms help ensure better tax collection.

Salt Lake City, for instance, is focused on reducing legal risks while improving service access. The AI tool helps reviewers by generating summaries of old PDFs, sometimes created decades ago by people no longer with the agency.

Open Source and Future Features

Built with Ruby on Rails, the AI PDF tool is open source and available on GitHub. Code for America continues to support Georgia and Salt Lake City with a small expert team through the deadline next April. The tool was designed for flexibility, allowing agencies to swap AI models as needed.

Future enhancements may include options to remediate files directly within the tool, such as marking PDFs for removal, converting them to HTML, or flagging them for further review. These capabilities would streamline the remediation process further.

Why This Matters

Government websites serve as the primary access point to public services. Despite cuts in federal programs and staffing, state and local governments remain responsible for providing equitable access. Fixing PDFs is a tangible step toward making government services more accessible and reducing barriers for people with disabilities.

With billions of PDF downloads occurring annually across federal agencies—many representing essential services like tax forms—ensuring these documents are accessible is critical. Tools like Code for America’s AI PDF manager provide practical support to help governments meet their legal obligations and serve all members of their communities effectively.

For government professionals interested in AI applications and improving accessibility, exploring practical AI tools can offer valuable assistance. Learning more about AI-driven solutions may prove beneficial in managing similar challenges in your agency’s digital services.