Accessibility in Video Streaming: What WCAG Actually Requires

TL;DR:

• WCAG 2.1 has five specific success criteria for video: captions for prerecorded (1.2.2, Level A) and live (1.2.4, Level AA) content, audio description or a media alternative for prerecorded video (1.2.3, Level A), and full audio description (1.2.5, Level AA).
• Auto-generated captions do not satisfy WCAG on their own; the standard requires captions that accurately represent both dialogue and meaningful non-speech sound, which most auto-caption engines still get wrong on jargon, punctuation, and sound cues.
• AIOZ Stream's Caption API covers the captions requirement directly, with a documented, verifiable field set. There is no dedicated audio-description feature in the API today, confirmed by checking the same model definition used to verify captions.
• Transcripts are not a named WCAG video criterion by themselves, but they are the most practical way to satisfy 1.2.1 (audio-only and video-only alternatives) and they double as an SEO asset, covered by AIOZ's separate Transcript API.
• US legal deadlines shifted in 2026: a DOJ interim rule pushed back Title II public-sector compliance dates to April 2027/2028, but private businesses are not covered by that extension and face active ADA litigation now.

What WCAG actually requires for video, criterion by criterion

"WCAG compliance for video" gets used as a vague catch-all, but the standard itself is specific. The relevant rules live under Guideline 1.2 (Time-based Media) in WCAG 2.1, and only five of them apply to typical on-demand or live video:

• 1.2.1 Audio-only and Video-only (Prerecorded), Level A. Requires a text alternative for content that is audio-only or video-only (no synchronized track to caption).
• 1.2.2 Captions (Prerecorded), Level A. Captions must be provided for the audio content of prerecorded synchronized media.
• 1.2.3 Audio Description or Media Alternative (Prerecorded), Level A. Either an audio description track or a full text alternative must be provided for prerecorded video content.
• 1.2.4 Captions (Live), Level AA. Captions must be provided for live audio content in synchronized media.
• 1.2.5 Audio Description (Prerecorded), Level AA. A dedicated audio description track is required for prerecorded video, not just the Level A fallback of a media alternative.

The distinction between 1.2.3 and 1.2.5 trips people up constantly: 1.2.3 (Level A) can be satisfied with either an audio description track or a full text transcript of the visual content, while 1.2.5 (Level AA) specifically requires the audio description track itself, no transcript substitute allowed. Most public-facing sites target Level AA, which means both captions and audio description end up mandatory, not optional add-ons.

Why auto-generated captions usually fail WCAG on their own

1.2.2 does not just require "some text under the video." It requires captions for all meaningful audio content, including non-speech sounds that carry information (a doorbell, applause, a phone ringing) when that sound is relevant to understanding the content. Automatic speech-to-text tools transcribe dialogue reasonably well for clean audio, but they routinely miss industry-specific terminology, mispunctuate in ways that change meaning, and skip non-speech cues entirely because they are only listening for speech. A caption file that passes a basic "captions exist" check can still fail an actual WCAG audit if it is riddled with these gaps. For anything beyond casual internal video, auto-captions need a human review pass before they can be called compliant, not just published as-is.

What AIOZ Stream actually supports today

Checking this against AIOZ Stream's own documented API rather than assuming coverage: the Caption API lets you attach a caption track to a video with a language code, a default flag, and a status field, and it is the mechanism that satisfies 1.2.2 and 1.2.4 directly, since captions are captions whether the source is auto-generated or professionally edited. The Go client's generated model confirms the exact field set: description, is_default, language, status, url, five fields total, nothing audio-description-specific among them, re-verified against the same model definition used for the earlier captions-vs-subtitles-vs-SDH breakdown.

That absence matters for 1.2.5 specifically. AIOZ Stream has no dedicated audio-description feature: no separate audio-track type, no flag distinguishing a description track from a regular caption or subtitle track. If a business needs Level AA audio description, the practical path today is producing the description as a separate audio file (or a mixed track with the description narration blended in) and hosting it as its own media asset, since the platform's caption mechanism is built for text tracks, not a second audio stream. This is the same kind of honest, verified gap already documented for DRM on AIOZ Stream: a real limitation worth knowing before it becomes a compliance surprise, not something to gloss over.

Transcripts are the more directly useful tool here in practice, even though WCAG does not name "transcript" as its own numbered criterion for standard video. A full transcript satisfies 1.2.1 for audio/video-only content, and it can help satisfy 1.2.3's "media alternative" option at Level A. AIOZ Stream's separate Transcript API generates and stores exactly this, and unlike audio description it maps cleanly onto AIOZ's existing text-track infrastructure.

The keyboard-accessible player requirement people forget

WCAG's video-specific criteria only cover the media itself. The player controls fall under general WCAG requirements, most importantly 2.1.1 (Keyboard), which requires that play, pause, volume, captions toggle, and seeking all work without a mouse. A video that has perfect captions and a full transcript still fails an accessibility audit if the play button can only be clicked, not reached by tabbing and activated with Enter or Space. This is a player-implementation detail, not a content-encoding one, so it depends on how the embed is built rather than anything the Media API controls directly. It is worth checking against AIOZ's own player documentation and against the W3C's own reference implementation for video player controls if keyboard accessibility specifically is a compliance requirement for a given deployment, since API-level caption support does not automatically make the surrounding player UI keyboard-operable. The W3C's own guidance on Understanding 2.1.1 (Keyboard) is the authoritative reference for exactly which interactions this covers.

The legal deadlines actually changed in 2026

The compliance timeline shifted this year in a way that is easy to miss if the only reference point is an older article. The Department of Justice issued an interim final rule in April 2026 that pushed back Title II compliance deadlines for public entities: April 26, 2027 for entities serving populations of 50,000 or more, and April 26, 2028 for smaller ones. The Department of Health and Human Services followed in May 2026 with a matching extension for organizations covered under Section 504. Those extensions apply specifically to public entities and federally funded programs.

Private businesses are a different story. There is no equivalent extension for them, and ADA Title III litigation over inaccessible video content continues actively regardless of the public-sector timeline shift. Section508.gov's own guidance on synchronized media still lists the same five WCAG success criteria described above as the baseline federal accessibility standard for video, extension or no extension, which is a useful cross-check that the underlying technical bar has not moved even though some deadlines have. A business assuming "the deadline got pushed back" based on the 2026 DOJ news without checking whether that extension actually applies to them is exactly the kind of mistake that leads to a lawsuit rather than a compliance grace period, and it is worth confirming which category a given organization falls into directly rather than assuming the extension is universal.

FAQ

Does WCAG require captions for every video, no exceptions?
No. 1.2.2 applies to prerecorded synchronized media specifically. Truly audio-only or video-only content falls under 1.2.1 instead, which requires a text alternative rather than captions in the traditional sense.

Is a transcript enough to satisfy WCAG on its own?
For 1.2.1 (audio/video-only content), yes. For standard video with both a visual and audio track, a transcript does not replace the caption requirement (1.2.2/1.2.4), and it only satisfies 1.2.3's audio-description alternative at Level A, not the stricter Level AA requirement in 1.2.5.

Can AIOZ Stream generate WCAG-compliant captions automatically?
The Caption API stores and serves caption tracks regardless of how they were produced. Whether an auto-generated caption file meets the WCAG accuracy bar depends on the transcription quality and whether it was reviewed, not on the API itself.

Does AIOZ Stream support audio description tracks?
Not as a distinct feature. The documented Caption API has no field distinguishing an audio-description track from a text caption track, confirmed directly against the API's model definition. A description track would need to be produced and hosted separately today.

What is the difference between WCAG Level A and Level AA for video?
Level A is the baseline: captions for prerecorded content (1.2.2) and audio description or a media alternative for prerecorded video (1.2.3). Level AA adds captions for live content (1.2.4) and upgrades the audio-description requirement to a real audio track rather than just an alternative (1.2.5). Most organizations target AA as their compliance floor.

Did the 2026 ADA deadline changes apply to my business?
Only if it is a public entity or a Section 504-covered program. The DOJ's April 2026 interim rule and HHS's matching May 2026 extension both apply to public-sector and federally funded organizations specifically; private businesses remain subject to ongoing ADA Title III enforcement with no equivalent extension.

Related Reading

• Multi-Language Captions and Transcripts: A Developer's Guide
• Captions vs Subtitles vs SDH: What's Actually the Difference
• Media Chapter vs Media Transcript API
• Video SEO: How to Get Your Videos Indexed and Ranked by Google
• AIOZ Stream Video Player: Features and Customization Guide