AI guided accessibility recommendations

They are AI-generated ideas for reviewing or addressing a selected scan finding. Recommendations may include suggested wording, explanations, or examples, depending on the issue and available context.

Run a scan, open a supported finding, and use the available AI assistance in ScrollAccess to request guidance or a suggestion.

They are based on the finding details and context available to the assistant, but they may not include everything about your page, design, visitors, or business rules. Review each recommendation in its actual context.

No. Recommendations are suggestions. Where a supported Designer change is offered, review and confirm it before it is applied. Publishing remains a separate action in Webflow.

No. Suggestions are available only where the feature has enough relevant information. Some issues require design decisions, user testing, or manual expertise.

No. AI may misunderstand an image, miss surrounding context, or suggest a change that does not fit your site. Verify the suggestion before using it.

No. They provide guidance for individual findings and do not certify that a site meets WCAG or legal requirements.

Review the result in the page context, test the relevant experience, and publish through Webflow when you are ready. Then rescan where applicable. Some behavior still needs manual testing.

It may provide HTML or CSS examples when relevant. Treat code as a draft: inspect it, adapt it to your implementation, and test it before use.

No. The finding and its recommendation both need review. Inspect the affected element and determine whether the issue applies to your content and use case.

The assistant uses relevant finding details available in ScrollAccess and any text you provide. See the in-app privacy notice and ScrollAccess Privacy Policy for information about data transmission, retention, and AI processing.

Recommendations can help you decide what to investigate, but they do not replace manual testing. Check keyboard access, focus, screen-reader behavior, dynamic interactions, and full user journeys as appropriate.