Audio Transcription and Accessibility: Captions, Transcripts, and the Law
Accessibility is not a feature you add at the end. It is the question of whether people can use what you made at all. For audio and video, the answer turns on one thing: is there a text version? A deaf person cannot follow a podcast without a transcript. A blind person cannot read a chart without a text description. A person with a cognitive disability may need the words written down to follow a fast speaker. Transcription is the step that makes spoken content available to all of them.
This guide explains what transcription does for accessibility, what the law requires in the United States and beyond, and how to publish transcripts and captions that actually work — not just files that exist. You can transcribe audio with AudioToText and download the result as text, subtitles, or timestamped JSON.
Why Text Is the Foundation of Accessibility
Every accessibility tool works with text. A screen reader reads text aloud. A braille display shows text as dots. A search engine indexes text. A translation tool translates text. Audio, on its own, works with none of them. The moment spoken words become text, they become usable by every one of those tools — and by every person who depends on them.
That is why a transcript is more than a convenience. For a deaf viewer, the transcript or the captions are the content itself; without them, the video is silent pictures. For a person with an auditory processing disorder, the text is the version they can follow at their own pace, re-reading the parts they missed. For a non-native speaker, the text is the dictionary they can consult while they listen. One text file serves all three, and it serves them at the same time, without any extra work per person.
There is a practical point too: accessibility features help everyone, not only the people who need them. Captions help the person watching in a noisy room. A transcript helps the person searching for one fact in an hour-long recording. Search engines, which are effectively blind, need the text to find the content at all. When you make content accessible, you make it better for the whole audience.
Who Benefits From Transcripts and Captions
Deaf and hard-of-hearing people. About 15% of American adults report some trouble hearing, according to the National Institute on Deafness and Other Communication Disorders, part of the National Institutes of Health. For them, captions are not optional — they are the only way to follow spoken content. See NIDCD for the figures and the research behind them.
Blind and low-vision people. A blind person listening to a video needs the visual information described in words — and a transcript is the place where that description lives. Audio description, the spoken narration of visual action, is usually written from a transcript of the video's dialogue first.
People with cognitive and learning disabilities. Written text can be re-read, slowed down, and broken into pieces in a way that live speech cannot. A transcript lets a person with dyslexia, an attention disorder, or a memory difficulty study the material at their own pace.
Non-native speakers. Reading along while listening is one of the most effective ways to follow a second language. A transcript turns a fast, idiomatic speaker into text that can be looked up word by word.
Everyone else. People in noisy places, people in quiet places, people on slow connections, people who want to search, quote, or skim — all of them use the transcript, and none of them need to ask for it specially. That is the point of accessibility done well: it disappears into the product.
What the Law Requires
The legal picture varies by country and by context, but the direction is consistent: public content must be accessible, and text is the way it is done.
United States — the ADA. The Americans with Disabilities Act requires public accommodations, including businesses that serve the public online, to communicate effectively with people with disabilities. The Department of Justice has made clear that this includes websites and online video. Its guidance on effective communication explains what is expected; see ada.gov for the official guidance. Courts have repeatedly held that online-only businesses are covered, and settlements routinely require captions on video and transcripts for audio.
United States — Section 508. Federal agencies must make their electronic content accessible, under Section 508 of the Rehabilitation Act. That means captioned video and transcripts for audio, meeting the Web Content Accessibility Guidelines (WCAG). If you supply content to a federal agency, expect to be asked for captions and transcripts as a condition of the contract.
United States — the FCC. The Federal Communications Commission requires captions on television and on much online video, with rules about accuracy, synchronisation, and completeness. The FCC publishes the current captioning requirements for broadcast and internet video.
Europe — the European Accessibility Act. From 2025, the European Accessibility Act requires many digital products and services sold in the EU to be accessible, including audiovisual media services. The standard is the same in practice: captions on video, transcripts for audio.
WCAG. The Web Content Accessibility Guidelines, published by the World Wide Web Consortium, are the technical standard almost everywhere. WCAG 2.1 requires captions for pre-recorded video (Success Criterion 1.2.2) and a transcript or audio description for pre-recorded content (1.2.1, 1.2.3). Meeting WCAG AA is the target most organisations aim for, and it is the standard courts and regulators use as the benchmark.
The law is a floor, not a ceiling. Meeting it means providing captions and transcripts that are accurate, synchronised, and complete — not just files that exist. A caption file full of errors does not meet the standard, and neither does a transcript that is published a month after the video.
How to Publish Accessible Transcripts and Captions
Step 1: Transcribe everything you publish
Every audio and video file you publish should have a text version. No exceptions for short clips, internal videos, or "temporary" content — temporary content is usually the content people need most. Make transcription part of the publishing checklist, not a task for later. Transcribe with AudioToText, and keep the transcript with the media file from the start.
Step 2: Provide captions on the video, and a transcript beside it
Captions and transcripts do different jobs, and you need both. Captions are timed to the video and watched with it; they serve the person watching now. A transcript is a separate document that can be read, searched, and used with a screen reader; it serves the person who wants the whole text. Publish the captions on the player, and put the transcript on the page below it, as an HTML page or a linked text file — not as a PDF image, which a screen reader cannot read.
Step 3: Make the transcript easy to find and use
A transcript buried three clicks deep might as well not exist. Put a "Transcript" link directly under the player, visible without scrolling. Write the transcript as structured text with headings, not as one long block. Give it a descriptive title and the date. If the content has speakers, label them. If it has sound that matters — [music], [applause] — mark it. A screen reader user navigates by headings; a transcript without them is a wall of text.
Step 4: Check accuracy, synchronisation, and completeness
The FCC's caption quality rules are a good checklist for any caption file, whether or not the FCC regulates you: accurate (the words are right, including names and numbers), synchronous (the captions appear when the words are spoken), and complete (everything spoken is captioned, from start to finish). Check all three before you publish. Our guide to adding subtitles to video covers the practical steps, and our accuracy guide explains what drives transcription errors and how to catch them.
Step 5: Do not rely on auto-captions alone
Platform-generated captions are a starting point. They are not checked, they are often wrong on names and technical terms, and they are rarely complete — they drop words when the audio is unclear instead of marking them. Upload your own checked caption file instead. It takes twenty minutes for a ten-minute video, and it is the difference between accessible and merely captioned.
What Good Accessible Content Looks Like
| Element | What to do | Why it matters |
|---|---|---|
| Captions | Timed SRT or VTT file, checked, on the player | Deaf viewers follow the video as it plays |
| Transcript | Full text on the page, with headings and speaker labels | Screen readers, search, and reading at one's own pace |
| Sound notes | [music], [applause] where the sound matters | Deaf viewers know what hearing viewers hear |
| Language | Set the language on the media and the transcript | Screen readers pronounce the text correctly |
| Player | Keyboard-accessible, with a visible captions button | People who cannot use a mouse can still watch |
A page that has all five is accessible in practice, not just on paper. A page that has only the first is better than nothing, but it leaves out the readers, the searchers, and the screen reader users — which is to say, it leaves out most of the people accessibility is for.
Common Mistakes to Avoid
Publishing the video and adding captions "later". Later rarely comes, and the first viewers — the ones who share and link — never see the captions. Make the caption file part of the publish step.
Using a PDF scan as the transcript. A scanned image of text is not text. A screen reader cannot read it, a search engine cannot index it, and a braille display cannot show it. Publish the transcript as HTML or plain text.
Hiding the transcript behind a download. A transcript that must be downloaded before it can be read is a transcript most people will not read. Put the text on the page, and offer the download as well.
Forgetting the language setting. A screen reader that does not know the language will pronounce the words wrong. Set the language on the video, the captions, and the transcript page, every time.
Treating accessibility as a legal chore. The law sets a minimum. The audience sets the standard. A transcript that is accurate, complete, and easy to find serves the whole audience — and it brings search traffic, better retention, and a wider audience with it. See pricing to find the plan that fits your publishing volume — the Free plan covers 60 minutes a month, and Starter covers 10 hours.
Frequently Asked Questions
Do I need captions if my audience is small?
Yes. Accessibility does not depend on audience size. A small audience can include a deaf viewer, and that viewer has the same right to follow your content as any other. The cost is the same whether ten people watch or ten thousand.
Are transcripts enough on their own, without captions?
No, not for video. A transcript serves the reader; captions serve the watcher. A deaf viewer watching a video needs the words timed to the picture. Provide both.
What about live video?
Live captions are harder, but the same principle applies: provide them where you can, and publish a corrected transcript afterwards. For a live event, a corrected transcript published the same day is a reasonable standard.
How accurate must captions be?
As accurate as you can make them. The FCC's standard is accuracy, synchronisation, and completeness, and that is a good target for any content. Names, numbers, and the main points should be right; a few small errors in filler words are acceptable, but a wrong number is not.
Does accessibility help with search?
Yes — substantially. Search engines read transcripts and caption files, so accessible content ranks for the words spoken in it. Accessibility and search optimisation are the same work, done once.
Key Takeaways
- Text is the foundation of accessibility: screen readers, braille displays, search engines, and translation tools all work with text, not audio.
- Provide both captions on the video and a full transcript on the page, with headings, speaker labels, and sound notes.
- Check accuracy, synchronisation, and completeness before you publish — a file full of errors does not meet the standard.
- Do not rely on platform auto-captions; upload your own checked file.
- Accessibility serves the whole audience, and it brings search traffic with it.
!Accessibility: captions and transcript
Image description: The image should show a video player with captions on, and beside it three icons: an ear with a line through it, a screen reader symbol, and a magnifying glass over text. Arrows run from the video to each icon. A caption reads "One transcript serves every viewer — and every search engine."
References
- U.S. Department of Justice, Americans with Disabilities Act — effective communication: https://www.ada.gov/
- Federal Communications Commission — captioning requirements: https://www.fcc.gov/
- National Institute on Deafness and Other Communication Disorders: https://www.nidcd.nih.gov/
- U.S. Department of Education — accessibility in education: https://www.ed.gov/