Our target
Top AI Customer Service targets WCAG 2.2 Level AA in current standards-based browsers and with common keyboard and assistive-technology combinations. A target guides design and testing; it is not a claim that every page or control has achieved complete conformance.
Accessibility contact
- Accessibility contact
- accessibility@firstaiemployee.com
What we test
- Desktop and mobile navigation, skip links, focus order, and breadcrumbs.
- Headings, landmarks, disclosures, source notes, and long-form content.
- Tables and stacked alternatives without color-only meaning.
- The AI customer-service cost calculator, including labels, errors, progress, results, and reset.
- Contact, correction, privacy, and accessibility email links.
- Channel diagrams with text descriptions and reduced-motion behavior.
Navigation
Automated and keyboard-only checks cover the skip link, header navigation, mobile disclosure, breadcrumbs, table of contents, footer, focus order, focus visibility, and route transitions.
content
Checks cover landmarks, heading order, link purpose, ownership disclosures, evidence notes, readable measures, reflow, and instructions that do not depend on position or color alone.
tables
Tables require captions, scoped row and column headers, keyboard-reachable overflow, visible scroll cues, and a meaningful narrow-screen presentation. Status and winners use text or shape as well as color.
calculators
The AI customer-service cost calculator is checked for programmatic labels, keyboard operation, described errors, progress and status semantics, formula disclosure, results, reset, reflow, and color-independent meaning.
forms
The sites have no forms. Calculator fields and the contact email links are checked for labels, instructions, error association, keyboard operation, and focus handling.
charts
Charts and channel diagrams require a text summary, labeled states, and downloadable data where quantitative data is presented. Animation must honor the reduced-motion preference.
| Area | Method | Coverage | Recorded status |
|---|---|---|---|
| Automated accessibility | Playwright and axe-core rule checks in Desktop Chrome | All 55 canonical pages for critical and serious WCAG-tagged rule violations | automated check implemented |
| Keyboard | Playwright keyboard-only, non-assistive-technology task flows | navigation, one calculator, contact email links, focus order, focus visibility, and state changes at desktop and mobile reflow | automated check implemented |
| Zoom and reflow | Playwright viewport and document-overflow assertions | Representative pages at a 720 CSS-pixel viewport as the 200% reflow boundary, plus 320 CSS-pixel interaction flows | automated check implemented |
| Motion and contrast | Playwright reduced-motion, focus, semantic-state, and axe contrast rule checks | Reduced-motion behavior and representative interaction states, with axe rules across all canonical pages | automated check implemented |
Testing methods
Automated checks run in Desktop Chrome at 320, 768, and 1440 CSS-pixel viewports. Keyboard-only, non-assistive-technology checks cover navigation, one calculator, the contact email links, visible focus, semantic structure, state indicators, reduced motion, zoom, reflow, and contrast. Automated and keyboard results cannot establish full conformance.
Human-operated VoiceOver and NVDA smoke tests were expressly owner-waived for this release. They are not reported as passed and remain a manual assistive-technology evidence limitation, alongside production-field evidence that has not been collected.
Automated
Automated accessibility checks run in Desktop Chrome at 320, 768, and 1440 CSS-pixel viewports. They detect selected rule failures but cannot establish full WCAG conformance.
keyboard
Keyboard-only, non-assistive-technology flows cover navigation, one calculator, the contact email links, focus order, visible focus, disclosure controls, and meaningful state changes without pointer input.
screen reader
Human-operated VoiceOver and NVDA smoke tests were expressly owner-waived for this release. They are not reported as passed and remain an outstanding manual assistive-technology evidence limitation.
zoom/reflow
Non-assistive checks review 200% zoom and reflow at 320 CSS pixels for clipped text, lost controls, obscured focus, overlapping sticky elements, and two-dimensional scrolling outside data tables.
contrast
Automated and visual checks cover text contrast, component contrast, visible focus, state indicators, and content that remains understandable without color. Human judgment remains necessary.
Known limitations
| Recorded | Scope | Status | Reader impact and next evidence |
|---|---|---|---|
| Human-operated VoiceOver and NVDA smoke tests | not performed owner waiver | No pass result or complete screen-reader coverage claim is published. A future human assistive-technology review may add evidence without changing the current owner waiver record. | |
| Production-field accessibility evidence | not collected before production traffic | Local tests cannot prove every browser, assistive technology, network, or production delivery combination. Review confirmed barrier reports and production observations after an authorized release. |
Report a barrier
Email accessibility@firstaiemployee.com with the page URL, the task you were attempting, the barrier, browser or assistive technology when useful, and your preferred contact method. You may request the information in an alternative accessible format. Do not include unnecessary sensitive data.
Information to include
Email accessibility@firstaiemployee.com with the page URL, attempted task, barrier, and useful browser, device, or assistive-technology context. Do not send unnecessary sensitive data.
alternative format
A reader may request the relevant publication information in an alternative accessible format. The request should identify the page or content and a usable contact method, without unnecessary sensitive details.
Assessment and update date
This statement was reviewed on . It is reviewed after material navigation, contact, calculator, table, or design-system changes and after new manual assistive-technology findings. Confirmed barriers and material remediations are tracked in dated internal editorial records.