Fizz LabsAccessibility Check

ADA Title II: check the PDFs on your government website.

By Fizz Labs · Last checked: 2026-11-09

The Justice Department's rule requires state and local government web content, including documents, to meet WCAG 2.1 Level AA. Start by finding out what you have. This free check reads up to 50 PDFs on your site and reports which have tags, a document language and real text.

Not legal advice. Not a guarantee of compliance. This page summarizes public sources as of 2026-11-09. Whether the rule or an exception applies to your organization is a legal question for your counsel. Our check looks at a few technical signals in each PDF; it does not test whether a document is accessible.

Compliance dates

Compliance dates
Your public entityCompliance date (WCAG 2.1 AA)
Total population of 50,000 or moreApril 26, 2027
Total population under 50,000, and special district governmentsApril 26, 2028

These dates come from the interim final rule published in the Federal Register on April 20, 2026 (document 2026-07663), which moved each date by one year and left the rest of the 2024 rule unchanged. The rule asked for public comments until June 22, 2026; check the Federal Register for any later change. Plain-language summary of the rule: ADA.gov.

The five exceptions, with their conditions

Each exception has conditions. Content that misses one condition is not covered.

  1. Archived web content. Only if all four conditions hold: it was created before the compliance date (or reproduces older physical media), it is kept only for reference, research or recordkeeping, it is stored in a special area for archived content, and it has not been changed since it was archived.
  2. Preexisting conventional electronic documents. PDFs, word processing, presentation and spreadsheet files available before the compliance date. This does not cover documents people currently use to apply for, access or take part in your services, such as a permit application form, and an old document that is updated later loses the exception.
  3. Content posted by third parties. Content posted by members of the public or others not acting for you, and not posted under a contract, license or other arrangement with you.
  4. Individualized, password-protected documents. Documents about a specific person, property or account that are password-protected or otherwise secured.
  5. Preexisting social media posts. Posts your organization made before its compliance date.

The rule also has general limits (fundamental alteration, undue burden) that require a documented decision; ask counsel.

Our check does not decide which PDFs fall under an exception. The CSV gives you facts to decide with: where each file was found, the creation date stored in the PDF and the server’s Last-Modified date.

Free PDF check

We follow your robots.txt, read at most 50 PDFs, and request only the first and last 64 KB of each file. We don't store your site address or results (your public robots.txt may stay in our cache for up to 10 minutes).

Or paste up to 50 PDF links from this site, one per line

Reading the results

Tagged (review content)
The file has a tag tree, is marked as tagged, declares a language and has real text. A person still needs to check reading order, alt text, headings and tables.
Tagged, no document language
Tags exist, but the document language is not set, so a screen reader may read it with the wrong voice.
Tagged, not marked
A tag tree exists, but the file does not declare itself as tagged (MarkInfo).
No tags
The file has text but no tag structure. It usually needs tagging in the authoring tool or a PDF editor.
Likely scanned image (no text layer)
We found images but no fonts or text. Scanned pages need text recognition and tagging, or a replacement document.
Unknown (structure not found in sampled bytes)
The parts we read did not include the document catalog, so we cannot say. This is common for large files; it is not a failure.
Skipped: blocked by robots.txt
Your robots.txt does not allow our checker to read this file, so we did not request it.

Need every PDF, not just 50?

We generate an automated, software-generated PDF inventory of your full site (up to 2,000 files): a software-generated report with page counts and a summary that sorts files for your team's review. You decide what is exempt and what to fix first.

Target delivery: 5 business days after we confirm scope.

Online checkout is coming. Public-sector purchasing paperwork can be discussed by email.

Questions

Does a tagged PDF mean it's accessible?

No. Tags are necessary for most screen reader use but not sufficient; reading order, alt text, headings and tables still need a person to check.

Why only 50 PDFs?

The free check is a sample to show the size of the job.

Do you store anything?

We count that a check was run. We don't store your site address, the PDF links or the results; the CSV is built in your browser. Your site's public robots.txt may stay in our server cache for up to 10 minutes so repeat checks don't refetch it.

Can you fix our PDFs?

We don't convert or fix PDFs. We help you find and sort them.

Sources

Related