Captions That Count: What WCAG and the ADA Ask of Short-Form Clips

A Caption Track Is Not the Same as Accessible Captions

Most short-form captions are designed for retention: two to four words at a time, large, animated, timed to pull the eye through the clip. That is a legitimate job, and it is not the same job as access. A viewer who cannot hear the audio needs to know who is speaking, what was said in full, and what the sound is doing when nobody is talking.

The audience is not small. The World Health Organization reports that over 5% of the world's population, 430 million people, need rehabilitation to address disabling hearing loss (World Health Organization fact sheet). For public bodies in the United States, there is now a dated legal duty as well.

What WCAG Actually Asks of a Caption

The Web Content Accessibility Guidelines are the standard most laws point to. For prerecorded video, Success Criterion 1.2.2 sits at Level A, the minimum. The W3C's explanation of it is specific: "Captions not only include dialogue, but identify who is speaking and include non-speech information conveyed through sound, including meaningful sound effects" (W3C, Understanding SC 1.2.2). The same document lists a named failure, captions "omitting some dialogue or important sound effects."

Four points follow for clip makers:

• Burned-in captions can count. Captions "can be 'open captions' that are always displayed and cannot be turned off" (W3C Web Accessibility Initiative). The question is what they contain, not whether a viewer can toggle them.
• Automatic captions do not count until checked. "Automatically-generated captions do not meet user needs or accessibility requirements, unless they are confirmed to be fully accurate" (W3C WAI).
• Paraphrase is a failure. A retention style that drops filler words is fine. A style that rewrites what was said is not, because the criterion is the dialogue.
• Visual information is a separate criterion. At Level AA, Success Criterion 1.2.5 requires audio description for prerecorded video (W3C, Understanding SC 1.2.3). A talking-head clip where everything important is said aloud has little to describe. A clip whose point is a chart on screen does.

Who Is Legally on the Hook, and When

In the United States, the Department of Justice's ADA Title II rule covers state and local governments. "The Web Content Accessibility Guidelines (WCAG) Version 2.1, Level AA is the technical standard for state and local governments' web content and mobile apps," and the department's own summary names "requirements for captions for videos" as an example of what that standard asks (ADA.gov).

The dates moved this year. "On April 20, 2026, the Federal Register published the Department's Interim Final Rule (IFR) extending the compliance date for State and local government entities with a total population of 50,000 or more to April 26, 2027" (ADA.gov). Smaller governments and special district governments now have until April 26, 2028.

The rule lists exceptions, and one is for preexisting social media posts. Read it as what it is: relief for the back catalog, not for what a city, a public university or a transit agency posts once its date arrives. If you produce clips for a public body, or you are one, the clips you publish after that date are the ones to plan for now.

This is general information, not legal advice. Private businesses sit under different parts of the ADA and different case law, and a lawyer should look at your situation.

Retention Captions and Access Captions, Side by Side

The good news is that one caption track can do both jobs. The differences are additions, not replacements:

• Speaker identification. On a two-person podcast clip, a name or a color-and-name convention when the speaker changes. Color alone is not enough, because color is not available to everyone.
• Sound cues. [laughs], [music], [phone rings], when the sound carries meaning. A punchline that lands on a laugh needs the laugh.
• Complete dialogue. Every word that carries meaning appears, in order, even when the style shows three at a time.
• Sync. Words appear when they are spoken, not a beat later.
• Placement. Captions never sit under platform buttons or over a face. Our safe-zone guide has the published margins.

A Caption Spec for Every Clip

Hand this to whoever exports your clips:

1. Proofread the transcript before styling it. Our proofing pass takes five minutes.
2. Show every meaningful spoken word, in order, synchronized.
3. Name the speaker at each change when more than one person talks.
4. Add a bracketed sound cue wherever sound carries meaning.
5. Keep text inside the safe zone and off faces.
6. Keep contrast high: light text on a dark stroke or box holds up on any background.
7. Where the platform allows it, upload a caption file as well as burning text in, so viewers who need a different size or font can switch.

Where ClipForge Fits

ClipForge generates captions for each clip in three styles: Bold Pop, Karaoke and Minimal. The Free plan includes one style and Creator includes all three (pricing). Your brand kit sets caption size, position, case, lines and animation once for every export (Brand kit guide). Treat generated captions as a draft: proof them, and add speaker names and sound cues where a clip needs them. Accessibility is decided by what the captions say, and that part stays with you.

Keep Reading

• Your Auto-Captions Can Invent Words: How to Proof an AI Transcript
• Why Auto Captions Are Essential for Short-Form Video in 2026
• How to Add Animated Captions to Videos Without Editing Experience

, Rocky