WCAG 2.2 added nine success criteria, but only six of them affect you if you are targeting Level AA — four at AA and two at A, since AA conformance requires meeting every A criterion too. For an LMS those six land unusually hard, because they target exactly what learning interfaces do: drag-and-drop questions, small tap targets in quizzes, sticky headers that cover the focused element, and logins that ask you to solve a puzzle. The trap that catches most institutions is separate from the criteria themselves: your LMS can be fully AA-conformant while the courses running inside it are not. The platform and the content are audited separately, and procurement usually only asks about the platform.
Key takeaways
- Nine criteria are new in 2.2, but three are Level AAA. If you are targeting AA, six are in scope: 2.4.11, 2.5.7, 2.5.8 and 3.3.8 at AA, plus 3.2.6 and 3.3.7 at A.
- 2.5.7 Dragging Movements is the one that hits eLearning hardest. Every drag-and-drop question, every drag-to-reorder interaction, needs a single-pointer alternative that does not require dragging.
- 4.1.1 Parsing was removed. If your audit report still fails you on invalid HTML nesting under 4.1.1, it is auditing against 2.1 and should be re-scoped.
- Your LMS being AA-conformant says nothing about your courses. A SCORM package, an uploaded PDF and a third-party video are each separately in scope, and they are where real audits fail.
- Moodle targets WCAG 2.2 Level AA and was accredited across its main products in October 2025 — but a theme, a plugin or a badly built course can still take a conformant platform out of conformance.
- Automated tools find perhaps a third of issues. Keyboard-only navigation and a screen reader pass are where the rest live, and neither takes as long as people fear.
What actually changed in WCAG 2.2
WCAG 2.2 was published as a W3C Recommendation on 5 October 2023, with an updated version following on 12 December 2024; the 2023 version went on to become the ISO/IEC 40500:2025 international standard. It is backwards compatible — everything that conformed to 2.1 still conforms to the same criteria in 2.2 — so this is an addition to your obligations, not a rewrite of them.
Nine criteria are new. Here they are with their levels, because the level is what decides whether you have to care:
| Criterion | Level | In scope for AA? |
|---|---|---|
| 2.4.11 Focus Not Obscured (Minimum) | AA | Yes |
| 2.4.12 Focus Not Obscured (Enhanced) | AAA | No |
| 2.4.13 Focus Appearance | AAA | No |
| 2.5.7 Dragging Movements | AA | Yes |
| 2.5.8 Target Size (Minimum) | AA | Yes |
| 3.2.6 Consistent Help | A | Yes — AA includes all of A |
| 3.3.7 Redundant Entry | A | Yes — AA includes all of A |
| 3.3.8 Accessible Authentication (Minimum) | AA | Yes |
| 3.3.9 Accessible Authentication (Enhanced) | AAA | No |
And one criterion was removed. 4.1.1 Parsing is gone from WCAG 2.2. It required valid HTML nesting and unique IDs, and it was retired because modern browsers and assistive technology handle those cases without it. This matters practically: if a vendor's audit report fails you under 4.1.1, that report is written against 2.1 and needs re-scoping before you spend money fixing findings that are no longer requirements.
The six that hit an LMS hardest
Generic WCAG checklists treat these as abstract web requirements. In a learning platform they are not abstract at all — each one maps onto something your courses almost certainly do today.
2.5.7 Dragging Movements (AA) — the expensive one
Any function that works by dragging must also work with a single pointer without dragging, unless dragging is essential. In eLearning that is not a corner case, it is a whole question bank: drag-and-drop matching, drag onto image, drag-to-order sequencing, sortable lists, and the drag-to-reorder interactions built into most course editors.
What passes: a click-then-click alternative — select the item, then select the destination. What does not: a keyboard alternative alone. The criterion is about single-pointer operation, so a mouse user with limited dexterity, or someone using a head pointer, still needs a non-dragging path. Check every drag question type your authoring tool produces, because the tool decides this, not you.
2.5.8 Target Size (Minimum) (AA)
Targets must be at least 24 by 24 CSS pixels, with defined exceptions for spacing and inline targets. Quiz radio buttons and checkboxes are the usual failure, followed by the small icon rows on course pages, pagination controls, and the close button on modals. It bites hardest on mobile, which for frontline and blended-learning audiences is the primary device.
2.4.11 Focus Not Obscured (Minimum) (AA)
When an element receives keyboard focus, it must not be entirely hidden by author-created content. The classic offender is a sticky header — the theme pins a navigation bar to the top, the user tabs down a long course page, and the focused link scrolls underneath it, invisible. Sticky footers, cookie banners and floating chat widgets do the same thing from the other direction. This is a theme-level bug, and it is worth testing on your own theme specifically because it appears only when tabbing, which nobody does by accident.
3.3.8 Accessible Authentication (Minimum) (AA)
No cognitive function test in the login unless an alternative or a mechanism to assist exists. In practice: puzzle CAPTCHAs are out, transcribing distorted text is out, and "remember this pattern" is out. Allowing the browser or a password manager to paste into the password field is what usually brings a login into conformance — and blocking paste, still a common security habit, actively breaks it.
3.2.6 Consistent Help (A)
If help is available, it must appear in the same relative place across pages. LMS deployments fail this by accident: a support link in the footer of the dashboard, a floating chat widget in the course player, and a different contact route inside the quiz. Pick one location and keep it.
3.3.7 Redundant Entry (A)
Do not ask for the same information twice in one process. Multi-step enrolment forms, course sign-ups that re-ask for details the profile already holds, and assessment flows that collect the same identifiers at each stage all fail this. Auto-populate it or offer it for selection.
The gap that fails real audits: platform versus content
This is the single most useful thing on this page, and it is the thing procurement conversations routinely miss.
Three separate layers each carry their own conformance, and a claim about one says nothing about the others:
What this means in practice
- A vendor VPAT or ACR covers the shipped product — not your custom theme, not your third-party plugins, and certainly not your courses.
- An uploaded PDF is in scope. Scanned handouts with no text layer are the most common single failure in an institutional audit, and no LMS setting fixes them.
- Video needs captions and, for AA, audio description where visual information is not in the narration. Auto-generated captions are a starting point, not a pass — they mangle exactly the technical and proper nouns that carry the meaning.
- Timed quizzes engage 2.2.1 Timing Adjustable. If a time limit is not essential to what you are assessing, it must be adjustable or extendable. Most compliance quizzes have a timer for no defensible reason.
Moodle specifically
Moodle is in a stronger position here than most platforms, and it is still possible to take a conformant Moodle out of conformance without noticing.
Where Moodle stands
Moodle targets WCAG 2.2 Level AA, and its developer policy states the standard plainly: every component must be usable by all users, and accessibility is part of the design of every new feature rather than a later pass. Accreditation for 2.2 AA was achieved across its main products in October 2025. If you are choosing an LMS on accessibility grounds, that is a real and checkable answer, which is more than several commercial platforms can offer.
What your theme can undo
The platform being conformant does not make your site conformant, and the theme is where it usually goes wrong. Moodle's own developer guidance is specific about what themes must get right:
- Landmark regions must separate the page layout, each with a unique label and the correct ARIA role — implemented in the theme's layout files. A custom theme that drops these breaks screen-reader navigation site-wide.
- Visible focus indicators must survive your CSS. Removing focus outlines for aesthetic reasons is the single most common self-inflicted accessibility bug in any custom theme.
- Colour contrast is where brand palettes collide with the standard. A brand colour that fails contrast on white does not become acceptable because it is the brand colour — test it before it reaches a style guide.
- Bootstrap components are not accessible by default. Moodle's guidance says so explicitly and provides an aria JavaScript module to enhance them. A theme that pulls in stock Bootstrap markup without that enhancement ships inaccessible components while looking perfectly fine.
Sticky headers deserve a specific mention because of 2.4.11: many modern Moodle themes pin the navbar, and that is precisely the pattern that obscures the focused element when a keyboard user tabs down a long course page. If your theme has a sticky header, test it before your auditor does. Our guide to choosing a Moodle theme covers the wider trade-offs.
Auditing course content inside Moodle
Platform-level checks will not find a scanned PDF or a missing image description inside a course. Content-level accessibility checking is a separate capability — the Brickfield Accessibility Toolkit is the established option in the Moodle ecosystem for scanning course content at scale and reporting what needs fixing, and it exists precisely because the platform audit does not reach layer three.
The checklist
Grouped by where the work actually happens, because that is how it gets assigned. Everything here is Level A or AA — nothing in this list is AAA.
Platform and theme
| Check | Criterion |
|---|---|
| Every function reachable and operable by keyboard alone, with no traps | 2.1.1, 2.1.2 |
| Focus indicator visible everywhere, and never removed in CSS | 2.4.7 |
| Focused element never fully hidden by a sticky header, footer or banner | 2.4.11 |
| Text contrast 4.5:1; large text and UI components 3:1 | 1.4.3, 1.4.11 |
| Landmark regions present, uniquely labelled, correct ARIA roles | 1.3.1 |
| Every page has a unique, descriptive title | 2.4.2 |
| Usable at 200% zoom and at 320px width without horizontal scrolling | 1.4.4, 1.4.10 |
| Help link in the same relative position on every page | 3.2.6 |
| Login has no puzzle or memory test; paste into password permitted | 3.3.8 |
Courses and interactions
| Check | Criterion |
|---|---|
| Every drag interaction has a click-then-click alternative | 2.5.7 |
| Clickable targets at least 24×24 CSS px, or adequately spaced | 2.5.8 |
| Quiz timers adjustable or removable unless the limit is essential | 2.2.1 |
| Enrolment and assessment flows never ask for the same detail twice | 3.3.7 |
| Errors identified in text, not by colour alone, with a suggested fix | 3.3.1, 3.3.3, 1.4.1 |
| Every form field has a programmatically associated label | 3.3.2, 1.3.1 |
| Headings used in order and for structure, not for visual size | 1.3.1, 2.4.6 |
Media and documents
| Check | Criterion |
|---|---|
| Captions on all prerecorded video, human-corrected not raw auto-captions | 1.2.2 |
| Audio description where visuals carry information the narration does not | 1.2.5 |
| Meaningful images have alt text; decorative images are marked decorative | 1.1.1 |
| Uploaded PDFs are tagged and have a real text layer, not scans | 1.1.1, 1.3.1 |
| Nothing auto-plays audio; media has pause and volume controls | 1.4.2 |
- 1Scope the three layers separately, on paper
Platform, authoring output, content. Write down who owns each and what evidence exists for each today. Most organisations discover at this step that they have evidence for one layer and assumed it covered all three.
- 2Run an automated scan to clear the cheap findings
Contrast failures, missing alt text, unlabelled fields, heading order. Automated tools catch roughly a third of issues, which is a third you no longer have to find by hand. Do not mistake a clean automated report for conformance.
- 3Do a keyboard-only pass on your five most-used journeys
Unplug the mouse. Log in, find a course, open a page, take a quiz, submit it. This one exercise finds focus traps, invisible focus, sticky-header obstruction and unreachable controls — four criteria in about twenty minutes.
- 4Test one course with a screen reader, not the whole catalogue
NVDA on Windows or VoiceOver on macOS. One representative course tells you what is systematically wrong with how your courses are built, which is the finding that scales. Auditing everything first is how accessibility projects stall.
- 5Audit the interactions your authoring tool generates
Every drag-and-drop question type, tested for a non-dragging path, and every target measured against 24×24. This is a tool-level finding: if the tool cannot produce a conformant drag question, you need a different question type, not a different course.
- 6Sweep uploaded documents and video separately
Scanned PDFs and auto-captioned video are the two failures that no platform setting fixes and no vendor VPAT covers. In Moodle, a content-level accessibility toolkit will find them at scale.
- 7Write the accessibility statement from what you found, honestly
Name the standard and level, list known exceptions with a remediation date, and give a contact route. A statement admitting three known gaps with owners reads as competent; one claiming full conformance you cannot evidence is the one that creates liability.
Automated testing finds
- Colour contrast failures, reliably and at scale
- Missing alt attributes and unlabelled form fields
- Broken heading order and missing page landmarks
- Invalid ARIA usage and duplicate IDs
- Roughly a third of real issues — genuinely useful, never sufficient
Only manual testing finds
- Whether alt text is accurate rather than merely present
- Whether a drag interaction has a usable non-dragging alternative
- Whether focus is obscured by a sticky header when tabbing
- Whether the reading order and announcements actually make sense
- Whether a learner can complete the course, which is the only question that matters
Need something custom-built?
Remediating an existing catalogue and building accessibly from the start are different projects with different costs. We audit Moodle themes and course content against WCAG 2.2 AA, fix what is found, and build courses that pass at layer two and three rather than only at the platform. If your immediate need is an accessibility statement you can defend in a procurement round, we will tell you honestly what your current position supports.
Unplug your mouse for twenty minutes.
Log in, find a course, open a page, take a quiz and submit it — keyboard only. No tooling, no budget, no vendor. You will find focus traps, invisible focus indicators, controls you cannot reach and at least one sticky header swallowing the focused element. It is the highest-yield accessibility test in existence and the one almost nobody runs.
Frequently asked questions
How many new criteria does WCAG 2.2 add, and how many affect us at AA?
Nine are new. Three are Level AAA and do not apply if you are targeting AA, so six are in scope: 2.4.11 Focus Not Obscured, 2.5.7 Dragging Movements, 2.5.8 Target Size and 3.3.8 Accessible Authentication at AA, plus 3.2.6 Consistent Help and 3.3.7 Redundant Entry at A — because AA conformance requires meeting every Level A criterion too.
Is our LMS being WCAG 2.2 AA enough?
No, and this is the most common misunderstanding in LMS procurement. A vendor conformance claim covers the shipped platform. It does not cover your custom theme, your third-party plugins, the packages your authoring tool publishes, or the PDFs and videos in your courses. Those are separate audits, and they are where real audits fail.
What happened to 4.1.1 Parsing?
It was removed from WCAG 2.2 entirely. It required valid HTML nesting and unique IDs, and it was retired because browsers and assistive technology now handle those cases. If an audit report fails you under 4.1.1, it was written against WCAG 2.1 and should be re-scoped before you pay to fix findings that are no longer requirements.
Do drag-and-drop quiz questions still pass under 2.2?
Only if there is a way to complete them with a single pointer without dragging — typically select the item, then select the destination. A keyboard alternative on its own is not sufficient, because the criterion is about pointer operation. Whether your questions pass is usually decided by your authoring tool rather than by how you built the course, so test the tool's output directly.
Is Moodle accessible?
Moodle targets WCAG 2.2 Level AA and was accredited across its main products in October 2025, and its developer policy requires accessibility in the design of every feature. That is a genuinely strong position. It is also undone easily: a custom theme that removes focus outlines, omits landmark regions or ships stock Bootstrap components without Moodle's aria enhancements takes a conformant platform out of conformance.
Are auto-generated captions good enough?
Not for conformance. Automatic captions are a reasonable first draft and they consistently mangle proper nouns, technical terms and numbers — exactly the words carrying the meaning in training content. Budget for human correction. Where visuals convey information the narration does not, AA also requires audio description, which auto-captioning does not address at all.
Can a timed quiz be accessible?
Yes, but the timer engages 2.2.1 Timing Adjustable. Unless the time limit is essential to what you are assessing, learners must be able to turn it off, adjust it, or extend it. Most compliance quizzes carry a timer out of habit rather than assessment design — removing it is often the cheapest conformance fix available.
Where should we start if we have never done this?
The keyboard-only pass on your five most-used journeys. It costs twenty minutes, needs no tools or budget, and finds four separate criteria's worth of problems. Do that before commissioning an audit, because it tells you whether you are facing a few fixes or a rebuild.
Where to start
Do the keyboard-only pass this week. It is free, it takes twenty minutes, and it will tell you more about your actual position than a vendor's conformance claim will. Then scope the three layers separately and find out which of them you genuinely have evidence for.
Related reading:
- Choosing a Moodle theme — the trade-offs, including the theme-level decisions that affect conformance
- Course development services — building courses that pass at the content layer, not only the platform
- eLearning in Hindi, Tamil and Telugu — the neighbouring problem, where rendering and language rather than assistive technology decide whether the course is usable
- HRMS and LMS integration — relevant when accessibility evidence has to reach the employee record
If you need to know whether your Moodle and your courses would survive an audit, we will look at both layers and tell you plainly.
Or write to marketing@edzlms.com.