Use Case
Local and municipal government
Local and municipal authorities publish information people depend on every day. Forms, fees, eligibility rules, consultations, housing updates, planning documents, licensing details, service changes, and public notices all influence what residents understand, what they do next, and whether they can access the service they need. That information is no longer read only by people. Search engines, AI search, and AI agents now read the same material on their behalf. If the content is inaccessible, outdated, unclear, or trapped inside documents that cannot be properly interpreted, the problem extends beyond accessibility. AI readiness is not only about internal tools. It also depends on whether people and AI can reach, interpret, and trust what your organisation has already published. Much of that knowledge sits inside PDFs, often spread across websites, portals, archives, departments, and partner-controlled spaces. Left unresolved, this creates accessibility gaps, AI distortion risk, legal exposure, and human misrepresentation. It can leave residents working from the wrong information, AI systems repeating material without context, and the authority exposed to a version of its services it no longer fully controls.
Summary
Local and municipal authorities already carry live accessibility duties across public-facing content. The wider issue is that the same content now supports public trust, service accuracy, search visibility, and AI readiness. Much of that information sits inside PDFs spread across departments, archives, portals, partner bodies, and pages no central team can fully manage. As organisations grow, the exposure grows with them, because more published material can remain findable long after the details have changed. We help close that gap at 95% lower cost and 99% faster. It converts linked PDFs into inclusive web content automatically, without altering the original document. That helps residents, search engines, AI search, and AI agents access information in a form they can reach, interpret, and trust.
The Challenges
Scale, cost, and documents beyond reach
Discovery
41%Of Websites sit outside any central record
Government websites grow organically. Websites with documents have no central record, making the estate hard to map completely.
Duplication
19%Of documents appear in more than one location
Nearly a fifth of all PDFs are duplicated across sites. Manual approaches must locate and remediate each copy individually.
Scale
3,000+Hours to clear 1,000 eight-page PDFs
At 23 minutes per page, a modest estate of 1,000 PDFs requires over 3,000 hours of specialist time before any new content is considered.
The legal obligation is in force across multiple jurisdictions. In the United Kingdom, the Public Sector Bodies Accessibility Regulations have applied since 2018. In the United States, the DOJ ADA Title II final rule requires all entities serving 50,000 or more residents to meet digital accessibility standards by April 2026. In Australia, the Disability Discrimination Act and WCAG 2.1 apply to all government digital content. Across the European Union, the European Accessibility Act came into force in June 2025.
Government websites grow over time in ways that make a complete picture of the PDF estate difficult to establish. Data across government web estates shows that up to 41% of documents have no central record, and around 19% appear in more than one location.
Manual accessibility work requires a specialist to work through every page individually, adding structure, tags, reading order, and descriptions at an average of 23 minutes per page. Clearing 1,000 PDFs averaging 8 pages each requires over 3,000 hours of specialist time, before any new content is considered.
A proportion of PDFs on any government website sit outside the reach of manual remediation entirely. Documents produced by third-party organisations, formally adopted public records, files where the original source no longer exists, and content produced in platforms no longer in use all fall into this category.
The accessibility obligation applies to PDFs used for current services regardless of when they were first published, which means age alone does not determine whether a document needs to be addressed.
How It Works
How automation works for local government
Our approach keeps the work away from already stretched teams. We provide a secure tag, which is added to the relevant pages. ARTY then understands your visitors and their needs, providing inclusive content that follows government recommendations, with no further demand on your team.
The service discovers linked PDFs and converts each document into an accessible web page that sits alongside the original. The source file is not changed, removed, or replaced. The resident still has the original PDF available, while people and AI systems also get content in a form they can use.
This matters for third-party documents, formally adopted records, locked files, archived material, and content where no editable source exists. Local government does not need to build a manual remediation programme around every document, or maintain multiple versions across every department and service area.
The Outcome
Inclusion isn't just human
Residents can reach accessible versions of PDF content across the digital estate, regardless of who produced it, when it was published, or whether the original can be edited. The obligation is met without a capital programme, at 95% lower cost and 99% faster than manual remediation. Human access and AI access are both supported, while the authority stays in control of what is found, understood, and acted on, creating a clear advantage over those still leaving public information unmanaged.
The Detail
Real-world outcomes
| Authority | Estate | Manual cost | Manual time | Expand |
|---|---|---|---|---|
| US municipality 180,000 residents | 1,100 PDFs | ~$165,000 | ~3,300 hours | |
| UK council small district | ~1,000 PDFs | £96,000 to £120,000 | ~3,000 hours | |
| UK unitary authority large | ~5,800 PDFs | approaching £580,000 | ~17,400 hours |
Further Reading