Accessibility

OKO Labs builds software that brings students together to talk through mathematics. A product premised on every student having a voice in the conversation has an obligation to be usable by every student. Accessibility is a design requirement for us, not a compliance exercise.

Conformance target: WCAG 2.2 Level AA
Report date: August 26, 2026

Our commitment

Our conformance target. OKO targets conformance with the Web Content Accessibility Guidelines (WCAG) 2.2 at Level AA. WCAG 2.2 Level AA incorporates all Level A and AA success criteria from WCAG 2.0 and WCAG 2.1, including the WCAG 2.1 Level AA requirements referenced in state K-12 software accessibility statutes such as Colorado HB21-1110.

Where we stand today. OKO's application interface substantially conforms to WCAG 2.2 Level AA. Every student-facing and teacher-facing workflow can be completed using a keyboard alone and with a screen reader. Six success criteria are reported below as Partially Supports. Three of those six concern mathematics problem content that OKO sources from third-party high-quality instructional materials providers rather than the OKO interface itself; that content is under active audit and remediation. The other three are a layout reflow limitation at extreme browser zoom, one below-threshold contrast element in the page footer, and the absence of live captioning for student-to-student speech. Each is itemized with a target date in the Remediation Roadmap.

Beyond the standard. OKO's session experience is multimodal by default. Problem text and facilitator prompts are rendered on screen and read aloud through built-in text-to-speech, and students participate by speaking as well as by typing and clicking. This benefits students with reading difficulties, students with low vision, and multilingual learners. It is additive to the criteria below, not a substitute for them: built-in audio does not replace screen reader support, and OKO does not report it as such anywhere in this document.

Report a barrier

If you encounter an accessibility barrier in OKO, or need content in an alternate format, contact us at support@okolabs.ai with the word Accessibility in the subject line. We aim to acknowledge every report within three business days and to give a remediation assessment within ten. Reports from students, families, teachers, and district staff are all welcome, and no account or contract relationship is required to submit one.

Report information

Name of Product / VersionOKO (web application), as of August 2026
Report DateAugust 26, 2026
Evaluation DateJune 15–17, 2026
Product DescriptionOKO is an AI-powered mathematics learning platform for students in grades 3–12. Students join small-group sessions and solve state standard–aligned math problems through alternating phases of individual work and group discussion, facilitated in real time by an AI guide. Problem text and facilitator prompts are presented visually and read aloud through built-in text-to-speech. Teachers assign content by standard, set session length, and review reporting on student proficiency, collaboration, and progress. Accessed through a standard web browser, primarily on Chromebooks and other laptops.
Contact InformationGeneral accessibility inquiries: support@okolabs.ai. Report author: TJ Marston, Co-Founder and Chief Product Officer.

Notes

This report covers the OKO student session experience and the teacher-facing controls and reports, accessed via web browser on Chromebook and on macOS. It reflects the product as of August 2026, based on evaluation conducted June 15–17, 2026. Conformance is reported against WCAG 2.2 Level AA, which incorporates all Level A and AA success criteria from WCAG 2.0 and WCAG 2.1.

This is a self-evaluation. It was conducted by OKO Labs personnel familiar with the product's functionality and intended use, using the automated tools, manual keyboard testing, and assistive technology testing described under Evaluation Methods Used. No third-party accessibility audit has been commissioned to date. OKO has not inflated any conformance level in this report; where the evidence supported either reading, the lower level was recorded.

OKO's mathematics problem content draws on high-quality instructional materials (HQIM) aligned to state standards, sourced from multiple third-party providers. Conformance of the application interface has been evaluated directly. Problem content is under ongoing accessibility audit and remediation, and several content-level criteria are reported as Partially Supports to reflect variability across sourced items. Remediation is sequenced by usage volume, so the most frequently assigned content is addressed first. See the Remediation Roadmap.

OKO is not being evaluated at WCAG Level AAA. In accordance with the VPAT 2.5Rev Best Practices for Authors, the Level AAA criteria table has been removed from this report rather than left unanswered.

Evaluation methods

OKO was evaluated against WCAG 2.2 Level AA using a combination of automated tools and manual testing, including keyboard-only navigation and screen reader testing. Testing was performed by team members familiar with the product's functionality and intended use. Automated scans preceded manual testing so that mechanically detectable defects were remediated first and manual testing time was spent on issues only a human evaluator can surface.

Automated testing

  • Tools: Axe DevTools and Google Lighthouse.
  • Scope: Every screen in the teacher and student workflows listed below, including modal and error states triggered manually before scanning.
  • Evaluation date: June 15, 2026.

Manual testing

  • Methods: Keyboard-only navigation with pointing devices disconnected, followed by screen reader evaluation.
  • Assistive technology: ChromeVox screen reader on Chromebook using Google Chrome 148.0.7778.263; VoiceOver screen reader on macOS using Google Chrome 149.0.7827.201.
  • Note: OKO's built-in text-to-speech was disabled during screen reader testing so that screen reader output could be evaluated on its own.

Workflows evaluated: teacher experience

  • Educator sign-up (Google, Clever, ClassLink, Microsoft) and login
  • Account settings and onboarding
  • Main dashboard, including group codes and session setup
  • Creating and managing student groups and assignments; setting a group's learning focus
  • Selecting and searching standards for a group code; browsing curriculum standards and lesson mappings; viewing or changing the lesson or standard
  • Session history
  • Class Insights weekly performance reports; individual student activity and engagement; student proficiency and mastery; collaboration trends
  • Resource hub
  • Live observation mode for an in-progress group

Workflows evaluated: student experience

  • Landing page and group code entry, including authentication redirect
  • Session timeout screen
  • Interactive collaborative math session, including tech check (microphone, device), joining the session room, device settings, core consensus activity, in-session dashboard, lesson content and OKO assistant, collaboration feedback, and the closing celebration screen

Standards covered

This report covers the degree of conformance for the following accessibility standards and guidelines:

Standard / GuidelineIncluded in report
WCAG 2.0Level A (Yes) · Level AA (Yes) · Level AAA (No)
WCAG 2.1Level A (Yes) · Level AA (Yes) · Level AAA (No)
WCAG 2.2Level A (Yes) · Level AA (Yes) · Level AAA (No)

Conformance terms

These are the definitions published by the Information Technology Industry Council (ITI); OKO has not modified them.

TermDefinition
SupportsThe functionality has at least one method that meets the criterion without known defects, or meets with equivalent facilitation.
Partially SupportsSome functionality does not meet the criterion.
Does Not SupportThe majority of functionality does not meet the criterion.
Not ApplicableThe criterion is not relevant to the product.

Not Evaluated is not used in this report; it is permitted only for Level AAA, which is not covered here.

Results: WCAG Level A

When reporting on conformance with the WCAG 2.x success criteria, they are scoped for full pages, complete processes, and accessibility-supported ways of using technology, as documented in the WCAG 2.0 Conformance Requirements. Unless an exception is noted, each criterion was verified through automated testing (Axe, Lighthouse) and manual keyboard and screen reader testing (ChromeVox, VoiceOver) across the workflows above.

CriterionConformanceRemarks
1.1.1 Non-text ContentPartially SupportsInterface content provides appropriate text alternatives. Some sourced HQIM items include images (graphs, diagrams, equations) whose text alternatives are incomplete or vary by source. Under audit, prioritized by assignment volume. Roadmap item 1.
1.2.1 Audio-only and Video-only (Prerecorded)Not ApplicableNo prerecorded audio-only or video-only content. Session audio is live.
1.2.2 Captions (Prerecorded)Not ApplicableNo prerecorded synchronized media.
1.2.3 Audio Description or Media Alternative (Prerecorded)Not ApplicableNo prerecorded synchronized media.
1.3.1 Info and RelationshipsPartially SupportsInterface conveys structure programmatically (landmarks, heading hierarchy, form label association, data table headers). Some sourced HQIM presents tables or structured information as images. Under remediation. Roadmap item 1.
1.3.2 Meaningful SequenceSupportsReading and navigation order is programmatically determinable and matches visual presentation.
1.3.3 Sensory CharacteristicsSupportsInstructions do not rely solely on shape, color, size, location, orientation, or sound.
1.4.1 Use of ColorSupportsColor is not the only visual means of conveying information; session state, correctness, and collaboration indicators are paired with text or iconography.
1.4.2 Audio ControlSupportsBuilt-in text-to-speech is user-initiated and can be stopped. No audio auto-plays longer than three seconds.
2.1.1 KeyboardSupportsBoth workflows complete end to end by keyboard, verified with pointing devices disconnected.
2.1.2 No Keyboard TrapSupportsFocus can be moved away from every component, including modals, with standard keyboard commands.
2.1.4 Character Key ShortcutsSupportsNo single-character key shortcuts are implemented.
2.2.1 Timing AdjustableSupportsSession and phase durations are set by the teacher and are essential to the activity. Inactivity timeout is signaled before it takes effect.
2.2.2 Pause, Stop, HideSupportsNo content blinks or auto-updates in a way that cannot be paused, other than real-time session state essential to the activity.
2.3.1 Three Flashes or Below ThresholdSupportsNo content flashes more than three times per second.
2.4.1 Bypass BlocksSupportsSkip-to-main-content mechanism and semantic landmarks provided.
2.4.2 Page TitledSupportsEach screen has a unique, descriptive title.
2.4.3 Focus OrderSupportsFocus order preserves meaning and operability, including at phase transitions and when a new problem loads. Verified manually.
2.4.4 Link Purpose (In Context)SupportsLink purpose is determinable from link text or its programmatic context.
2.5.1 Pointer GesturesSupportsNo multipoint or path-based gestures required.
2.5.2 Pointer CancellationSupportsSingle-pointer functions complete on the up-event and can be aborted.
2.5.3 Label in NameSupportsAccessible names contain the visible text label.
2.5.4 Motion ActuationSupportsNo functionality is operated by device or user motion.
3.1.1 Language of PageSupportsDefault human language is programmatically set.
3.2.1 On FocusSupportsReceiving focus does not initiate a change of context.
3.2.2 On InputSupportsChanging a setting does not cause an unexpected change of context.
3.2.6 Consistent HelpSupportsHelp mechanisms appear in a consistent relative order where provided.
3.3.1 Error IdentificationSupportsInput errors are identified in text and described.
3.3.2 Labels or InstructionsSupportsLabels and instructions are provided; placeholder text is not used as a substitute for a programmatic label.
3.3.7 Redundant EntrySupportsPreviously entered information is auto-populated or available for selection within a process.
4.1.1 ParsingSupportsAnswered Supports for WCAG 2.0/2.1 per W3C guidance; removed as obsolete in 2.2.
4.1.2 Name, Role, ValueSupportsInteractive components expose name, role, state, and value, including custom components (standard picker, in-session answer controls). Verified with ChromeVox and VoiceOver.

Results: WCAG Level AA

Unless an exception is noted, each criterion was verified through automated testing (Axe, Lighthouse) and manual keyboard and screen reader testing (ChromeVox, VoiceOver) across the workflows above.

CriterionConformanceRemarks
1.2.4 Captions (Live)Partially SupportsAll AI facilitator speech is presented as on-screen text in addition to being spoken, so facilitator audio has a synchronous text equivalent. Live student-to-student speech is not captioned. Under evaluation. Roadmap item 4.
1.2.5 Audio Description (Prerecorded)Not ApplicableNo prerecorded synchronized media.
1.3.4 OrientationSupportsContent does not restrict its view to a single display orientation.
1.3.5 Identify Input PurposeSupportsInput purpose is programmatically identified for fields collecting information about the user.
1.4.3 Contrast (Minimum)Partially SupportsAll primary content, controls, fields, and interactive elements meet 4.5:1 (or 3:1 for large text). One element falls below threshold: the footer copyright notice over a decorative, variable-color background. No information is conveyed exclusively by it. Scheduled for correction. Roadmap item 3.
1.4.4 Resize textSupportsText resizes to 200 percent without assistive technology and without loss of content or functionality.
1.4.5 Images of TextPartially SupportsInterface text and HQIM prose are live text. Some sourced items present text as images. In the ongoing content audit. Roadmap item 1.
1.4.10 ReflowPartially SupportsMost content reflows to one column at high zoom. The multi-column teacher dashboard and fixed navigation may require two-dimensional scrolling at 400 percent. Data tables and complex graphs are exempt. Roadmap item 2.
1.4.11 Non-text ContrastSupportsComponent boundaries, icons, and focus indicators meet 3:1 against adjacent colors.
1.4.12 Text SpacingSupportsNo loss of content or functionality when the user overrides line, paragraph, letter, or word spacing to the specified values.
1.4.13 Content on Hover or FocusSupportsHover/focus content is dismissable, hoverable, and persistent.
2.4.5 Multiple WaysSupportsReport pages are reachable via persistent left-hand navigation and contextual links. Session-flow and setup pages are process steps and exempt.
2.4.6 Headings and LabelsSupportsHeadings and labels describe topic or purpose. Verified via heading-list navigation in ChromeVox and VoiceOver.
2.4.7 Focus VisibleSupportsA visible focus indicator is present on all keyboard-operable components.
2.4.11 Focus Not Obscured (Minimum)SupportsThe focused component is not entirely hidden by author-created content, including during modal and overlay states.
2.5.7 Dragging MovementsSupportsNo functionality requires dragging; single-pointer and keyboard alternatives exist.
2.5.8 Target Size (Minimum)SupportsPointer targets meet the 24-by-24 CSS pixel minimum, or qualify under the spacing or inline exceptions.
3.1.2 Language of PartsSupportsPassages in a non-default language are programmatically identified.
3.2.3 Consistent NavigationSupportsRepeated navigation occurs in the same relative order across screens.
3.2.4 Consistent IdentificationSupportsComponents with the same function are identified consistently.
3.3.3 Error SuggestionSupportsWhere an error is detected and a correction is known, it is suggested.
3.3.4 Error Prevention (Legal, Financial, Data)Not ApplicableNo legal commitments, financial transactions, or user-modifiable test-data submissions.
3.3.8 Accessible Authentication (Minimum)SupportsNo cognitive function test required. Educators use third-party SSO; students join with a group code.
4.1.3 Status MessagesSupportsStatus messages and dynamic session updates use ARIA live regions, announced without a change of focus, alongside visual and audio presentation. Verified separately in ChromeVox and VoiceOver.

Remediation roadmap

Six criteria are reported as Partially Supports. Target windows are stated by calendar quarter; progress is reported in each reissue.

#Item and affected criteriaPlanned workTarget
1Third-party HQIM problem content (1.1.1, 1.3.1, 1.4.5)Audit sourced items for missing text alternatives, structured information presented as images, and text rendered as images. Remediate where OKO controls rendering; raise defects with the provider where it does not. Sequenced by assignment volume.Highest-volume content by Q4 2026; remaining library through Q3 2027
2Reflow at 400 percent zoom (1.4.10)Rework the multi-column teacher dashboard and fixed navigation to collapse to one column at high zoom without two-dimensional scrolling.Q1 2027
3Footer contrast (1.4.3)Correct the contrast ratio of the footer copyright notice against its decorative background.Q4 2026
4Live captioning of student speech (1.2.4)Evaluate automated live captioning for student-to-student speech, including accuracy on child speech and math vocabulary and student-privacy implications. Scope an approach based on that evaluation.Evaluation complete Q2 2027

Keeping this current

  • Automated checks in the release pipeline. Accessibility scanning runs against every build, so mechanically detectable regressions are caught before release rather than at the next audit.
  • Quarterly manual passes. Keyboard and screen reader testing is repeated each quarter against new and changed feature areas, using the same methodology described in this report.
  • Annual reissue. This report is reviewed and reissued at least annually, and sooner when a significant product change or a roadmap completion warrants it.
  • Third-party audit. OKO intends to commission an independent accessibility audit to validate this self-evaluation. This report will be reissued to reflect its findings.

Revision history

VersionDateDescription
1.0August 26, 2026First published Accessibility Conformance Report, based on evaluation conducted June 15–17, 2026. Adds the Accessibility Statement, the self-evaluation disclosure, and the Remediation Roadmap. Supersedes all internal drafts, which were not published.

Legal disclaimer

This Accessibility Conformance Report describes the results of OKO Labs, Inc.'s self-evaluation of the OKO web application against the standards above, as of the report date. It is provided for informational purposes to assist customers and prospective customers in making preliminary assessments of the product's accessibility. It reflects OKO's good-faith assessment based on the evaluation methods described and is not a warranty, guarantee, or certification of conformance. Results may vary depending on the browser, operating system, assistive technology, and configuration in use, and on third-party instructional content assigned within the product. Target dates in the Remediation Roadmap are planning estimates, not contractual commitments. OKO Labs, Inc. will provide additional information about any item on request.

Voluntary Product Accessibility Template and VPAT are registered service marks of the Information Technology Industry Council (ITI).