What Accessibility Scores Establish — and Why Review Records Matter

A scanner score summarizes the checks a tool implements. A review record explains the actions taken on a particular file. Both can support an accessibility program, but neither establishes conformance or guarantees a favorable investigation outcome.

Read a score with its denominator

Ask which criteria, formats and versions the scanner evaluates, what it leaves untested, and whether the result describes the actual saved output. A score of 100 means the scoring method reported no penalized findings within its scope; it does not mean every applicable WCAG criterion passed.

There is no universal percentage of WCAG failures that every scanner detects. Coverage depends on the rules, input and evaluation method. Presence of alternative text does not establish that it accurately describes a teaching figure. A structural check alone cannot determine every reading-order or caption-quality issue.

Preserve the evidence behind a change

Useful review records include:

• The document identifier, source version and source/output checksums.
• The tool version, configuration, scan date, findings and untested areas.
• The proposed or applied change and its relationship to the original finding.
• The reviewer or responsible actor, decision, timestamp and reason.
• Validation and manual-check results for the saved output.
• Unresolved work, ownership, agreed follow-up and retention controls.

Distinguish automated actions from human decisions. Do not invent per-finding attribution when a run recorded only aggregate outcomes. A high acceptance rate is not by itself evidence of rubber stamping; assess the substance and completeness of the review.

Procurement questions

Ask vendors to demonstrate export and retrieval of records for an exact document version. Machine-readable records can support analysis; a readable report can support human review. Check access controls, integrity, retention and correction history rather than insisting that one file format is the only acceptable evidence.

Records should follow institutional policy, applicable law and any case-specific obligations. FERPA does not impose a universal retention period for every accessibility trail. Counsel should determine relevant requirements for student records and legal holds.

Keep records alongside substantive accessibility work

The DOJ Title II rule sets WCAG 2.1 Level AA requirements for covered state and local government web content and mobile apps, subject to the rule's scope and exceptions. The April 2026 IFR extends the dates to April 26, 2027 for public entities with a population of 50,000 or more and April 26, 2028 for smaller public entities and special district governments. Determine the institution's cohort using DOJ's population rules; enrollment and federal funding alone do not determine it. Existing accessibility obligations continue before those dates.

A policy, coordinator and documented process can support governance, but documentation is not a substitute for accessible content. Resolution agreements address specific parties and facts; they do not establish one universal eight-field export requirement for every institution.

Aelira Core is beta software. Supported workflows can apply eligible partial fixes and record outcomes, but some inputs need manual work or cannot produce a downloadable remediated file. Review the actual saved output, unresolved findings and content preservation before publication. Automated scores describe implemented checks rather than accessibility conformance. Inspect the public source and release documentation and evaluate a representative authorized sample before planning a rollout.

Further reading

These links provide additional context; each source should be assessed for its own scope.

AI-Assisted University Accessibility: Review, Evidence and Legal Scope

How to Use OCR Resolution Agreements in Accessibility Planning

Resource 3: www.w3.org

Reviewing AI-Generated Alt Text for Academic Figures