WCAG audio description
for corporate video,
in plain language.

Short answer: if your brand's website hosts video with audio, WCAG 2.1 Level AA expects (1) a synchronized caption file on every prerecorded video, (2) live captions on every live event, and (3) a full audio description audio track on every prerecorded video where the visuals carry information the soundtrack does not explain. A text transcript meets the Level A bar; it does not meet the Level AA bar.

WCAG (Web Content Accessibility Guidelines) is a W3C standard. Version 2.1 has 50 success criteria at Level AA, the conformance level US courts apply in ADA Title III digital cases and the level the Department of Justice formally adopted for ADA Title II in 2024. For video, five criteria sit under Guideline 1.2 ("Time-based Media"). Four of them matter for corporate work.

What does WCAG 2.1 AA actually require for corporate video?

Four criteria do almost all of the work. Three are Level A (minimum conformance); two are Level AA (the level courts apply). At AA, both 1.2.4 (live captions) and 1.2.5 (audio description) become binding in addition to the A-level requirements.

What does audio description mean, exactly?

Audio description (AD) is a narration track that describes important visual information that cannot be understood from the original soundtrack. W3C's exact definition: "narration added to the soundtrack to describe important visual details that cannot be understood from the main soundtrack alone." AD is an audio track, not a text file. The W3C's Understanding doc clarifies the narration covers actions, characters, scene changes, and on-screen text, inserted during natural pauses in dialogue.

Why this matters for brands: at Level A (criterion 1.2.3), a brand can satisfy the obligation with either a separate AD audio track or a full text alternative for the media. At Level AA (criterion 1.2.5), the text-alternative option is removed. Audio description is mandatory in audio form. This is the criterion most brands inadvertently fail — they assume their transcript covers them, and at AA, it does not.

Which of your videos need audio description?

Any prerecorded video where the visuals carry information the soundtrack does not already speak aloud. That is most of them — product demos, brand films, recruitment videos, training videos, anything with on-screen text the VO does not read.

Worked examples for a typical corporate video catalog:

Homepage hero loop, silent, no on-screen text: covered by 1.2.1 (video-only prerecorded). Usually satisfied by a near-by text label or description.

Homepage hero loop, silent, with on-screen text: the on-screen text needs a text equivalent if the audio carries no description of it. If there is any audio at all, treat as synchronized media; 1.2.2 + 1.2.5 apply.

Product demo with VO: 1.2.2 (captions) plus 1.2.5 (AD). Both required at AA.

Webinar or town hall (live): 1.2.4 (live captions). Once the recording is posted as an on-demand replay, 1.2.2 and 1.2.5 apply to the recorded version.

Recruitment, culture, or brand film: 1.2.2 (captions) plus 1.2.5 (AD). Brand films almost always have visual storytelling the VO does not narrate; AD is the default obligation here.

Investor relations / earnings replay: once a live earnings call becomes a posted replay, 1.2.2 and 1.2.5 apply to the recorded version.

Is a transcript enough at Level AA?

No. At Level A (criterion 1.2.3), a full text alternative for the media satisfies the requirement. At Level AA (criterion 1.2.5), it does not. The W3C's Understanding doc states the gap directly: "If they wish to conform at Level AA, authors must provide an audio description — a requirement already met if they chose that alternative for 1.2.3, otherwise an additional requirement."

This is the most common compliance miss. Brand teams ship a video, attach a transcript file, and stamp the deliverable compliant. They are correct at A. They are non-compliant at AA. And AA is the level federal courts apply in ADA Title III digital cases and the level DOJ adopted for Title II. The transcript covers 1.2.3 at A; it does not cover 1.2.5 at AA.

Live webinars: what changes when the recording goes on-demand?

While a webinar is live, criterion 1.2.4 applies: real-time captions are required. Most enterprise webinar platforms now offer live captioning either via human stenographer or via ASR (automatic speech recognition). The FCC's quality standard for live programming is the broadcast pillars — accurate, synchronous, complete, appropriately placed — though WCAG itself does not specify a numeric accuracy floor.

Once the live event is posted as a recording, the obligation shifts. The recording is now "prerecorded synchronized media," which triggers 1.2.2 (captions, A) and 1.2.5 (audio description, AA). Brands often forget the second half. The recorded webinar that was acceptable while live is no longer acceptable at AA once it becomes the on-demand replay — it needs an AD pass before the link is shared.

The 10-item checklist for a corporate video team

Every video on the public site or app should clear these ten yes-or-no questions. Three or four "no" answers is normal for a back catalog; the floor for new content should be all ten "yes."

1. Is there a caption file (VTT, SRT, TTML, or SCC) attached to every prerecorded video with audio?
2. Do the captions include speaker identification and non-dialogue audio (sound effects, music cues) where meaning depends on them?
3. Do the captions cover any on-screen text the soundtrack does not already speak aloud?
4. Is there a separate audio description audio track on every prerecorded video where visuals carry information the main soundtrack does not?
5. If the video is live (webinar, town hall, earnings call), are real-time captions provided?
6. Once the live event is posted as a replay, does it have full prerecorded captions and AD?
7. Is the video player keyboard-operable — play, pause, volume, captions toggle reachable via Tab?
8. Does the page avoid autoplay-with-audio, or provide a visible pause control within 3 seconds (WCAG 1.4.2)?
9. For silent hero loops, is there a text equivalent describing what the video shows?
10. Are the captions accurate — reviewed by a human pass, not raw auto-generated YouTube output?

Three things to do this quarter

1. Audit the catalog against the checklist. Pull a list of every video hosted on or embedded in your domains and apps. Flag each on the 10 questions above. The output is your remediation backlog — ranked, dated, owned.
2. Set a delivery floor for everything new. Update your video team's Definition of Done to include all 10 items. No video ships without it. New work being compliant by default lets you spend remediation capacity on the back catalog instead of the front of the pipeline.
3. Prioritize remediation by traffic and criterion severity. 1.2.5 (AD) remediation requires production work — script, voice talent, mix — that 1.2.2 (caption correction) does not. Pages with high traffic and the AD gap go to the front of the queue. Document the rationale; a court or plaintiff's counsel reviewing a remediation plan will ask whether it was systematic.

Editor's note. Prepared by The Slate, Editorial. Published 11 May 2026. WCAG 2.1 criterion language is verbatim from the W3C Recommendation. Litigation statistics from Seyfarth Shaw's annual ADA Title III tracking reports. WebAIM Million figure from the 2026 study. Not legal advice. Consult counsel for entity-specific obligations.