Back to resources

Resource article

Auto-generated transcripts vs ADA-compliant closed captions

Why machine-generated text can help draft workflows but does not automatically satisfy accessibility obligations.

Last updated: 5/11/2026

Direct answer

Auto-generated transcripts are a useful starting point, but ADA-aligned captioning requires accurate, synchronized captions that include meaningful non-speech information.[1][2]

What this means in practice

  • WCAG 2.1 SC 1.2.2 requires captions for prerecorded synchronized media, with limited exceptions.[1]

  • Caption requirements include spoken dialogue and meaningful non-speech sounds, not just word-for-word transcript text.[1]

  • YouTube documents that automatic captions can misrepresent speech and should be reviewed and edited.[2]

  • A compliant workflow typically includes human quality checks for accuracy, timing, speaker context, and formatting.[1][2]

FAQ

Are transcripts and captions the same thing?

No. Captions are synchronized to media playback and include non-speech audio context. A plain transcript alone does not provide that synchronized experience.[1]

Can we start from automatic captions?

Yes, but YouTube itself recommends reviewing and correcting auto-captions because machine output may be inaccurate.[2]

Annotated sources

  1. [1] W3C WCAG 2.1 Understanding SC 1.2.2 (Captions, prerecorded)

    W3C WAI | WCAG 2.1 guidance

    Defines what captions must include and why synchronized captioning is required for accessibility.

  2. [2] YouTube Help: Use automatic captioning

    YouTube Help | Help article

    States auto-captions may misrepresent speech and should be reviewed/edited.

This article is informational and not legal advice. Organizations should consult counsel for legal determinations.