What Is Closed Captioning? Definition, Legal Requirements, and Practical Guide (2026)

Broadcast TV (CEA-608 / CEA-708). On US analog television, closed captions were encoded in Line 21 of the vertical blanking interval per the CEA-608 standard, decoded by the TV's built-in caption chip. Digital ATSC television uses the successor CEA-708 standard, which supports multiple caption tracks, richer styling, and up to 8 language streams. Broadcast captioning workflows produce SCC files that carry the raw CEA-608 byte pairs.

Web video (sidecar files). HTML5 video and modern streaming players read a separate caption file — most commonly SRT or WebVTT — and render the text as an overlay synchronized to the video timeline. Netflix and enterprise workflows use TTML (in the IMSC 1.2 profile) for stricter styling and multi-language delivery. YouTube accepts SRT, VTT, SBV, SCC, TTML, and DFXP.

Embedded vs sidecar. Captions can be muxed into container formats (MP4, MOV, MKV) as a text track alongside video and audio streams, or shipped as a separate file. Sidecar files are easier to update and localize; embedded tracks are easier to distribute in a single asset. Both remain toggleable at the player.

Live captions. For live-streamed video, captions are generated in real time by a human stenographer (CART), by automatic speech recognition (Google, Microsoft, Otter, Zoom), or by a hybrid pipeline. Live captions are transmitted over WebSocket, SRT protocol, or an in-band caption channel. Industry norm for compliance-grade live captions is under 5 seconds of end-to-end latency.

For a deeper look at the two most common web formats, see what is an SRT file and what is a VTT file.

The Department of Justice published its final rule on the accessibility of web information and services of state and local governments in the Federal Register on April 24, 2024 (89 FR 31320). The rule adopts WCAG 2.1 Level AA as the technical standard for public web content and mobile apps. Under WCAG 2.1 Level AA, both SC 1.2.2 (prerecorded captions) and SC 1.2.4 (live captions) are required.

Compliance dates are staggered by jurisdiction size: April 24, 2026 for large local governments (populations of 50,000 or more) and state government entities, and April 24, 2027 for small local governments (populations under 50,000) and special-purpose districts.

Primary source: Federal Register 89 FR 31320 · ada.gov/resources/web-guidance

ADA Title III (42 U.S.C. §12181 et seq.) prohibits discrimination in places of public accommodation. Since NAD v. Netflix (D. Mass. 2012, 869 F. Supp. 2d 196), courts have consistently applied Title III to web-only services and to the web presence of physical businesses (Robles v. Domino's, 9th Cir. 2019). The DOJ has not fixed a specific technical standard by regulation for Title III, but WCAG 2.1 Level AA is the de facto standard cited in enforcement and private lawsuits.

Primary source: ada.gov/topics/title-iii · DOJ web accessibility guidance

Section 508 of the Rehabilitation Act (29 U.S.C. §794d), as revised in the 2018 ICT Refresh, requires federal agencies to make their electronic and information technology accessible. The Revised 508 Standards incorporate WCAG 2.0 Level AA by reference and apply to procurement, development, and maintenance of ICT. Captions are required for all prerecorded and live synchronized media produced or funded by federal agencies.

Primary source: section508.gov · US Access Board ICT Standards

The CVAA (Public Law 111-260, October 2010) extended captioning requirements to IP-delivered video that was previously aired on US broadcast television. The FCC implemented CVAA captioning obligations in a phased schedule from 2012 through 2016, with distinct rules for prerecorded, live, near-live, edited-for-IP, and video clip categories.

Primary source: fcc.gov/consumers/guides/twenty-first-century-communications-video-accessibility-act-cvaa

The FCC captioning rules for television (47 CFR §79.1 et seq.) require broadcasters, cable operators, and satellite TV providers to caption almost all English- and Spanish-language programming. The rules also set four quality standards: accuracy (verbatim within the limits of the medium), synchronicity, completeness, and placement. These quality standards are the most widely cited benchmark for editorial review of any caption file, even outside the FCC's direct jurisdiction.

Primary source: fcc.gov/consumers/guides/captioning-television-and-video

California's Unruh Civil Rights Act (Civil Code §51 et seq., with recent amendments including AB 1836) sets a $4,000 per-violation minimum in private civil actions and has been the vehicle for a large share of state-level web-accessibility captioning lawsuits. New York's Human Rights Law (§290 et seq.) plus the New York City Human Rights Law generate a similar volume of private enforcement in the Southern and Eastern Districts of New York. Both statute frameworks apply WCAG 2.1 AA as the de facto conformance target through case law.

The European Accessibility Act (Directive (EU) 2019/882) has been in force across all EU Member States since June 28, 2025. It requires accessibility — including captions on audiovisual content and video communication — for a defined list of private-sector products and services placed on the EU market. Covered categories include e-commerce, consumer banking services, e-books and dedicated software, audiovisual media services (including their electronic programme guides), passenger transport services, ATMs and payment terminals, and smartphones and their operating systems.

Transition window. Products and services already on the market before June 28, 2025 have until June 28, 2030 under the Directive's transitional provisions. Some service contracts concluded before June 28, 2025 may run to their natural end no later than June 28, 2030.

Technical standard. The harmonized standard EN 301 549 is the primary conformance path for the EAA and incorporates WCAG success criteria for captions on prerecorded and live synchronized media. Enforcement is by Member State market-surveillance authorities and includes penalties, product withdrawal, and end-user complaint mechanisms.

Primary source: EUR-Lex Directive 2019/882 · European Commission — EAA

Even outside a regulatory obligation, three practical reasons make captions the default choice for published video.

Sound-off consumption is the norm on social feeds. Widely cited industry estimates put the share of social-feed video watched with sound off at around 85% (Digiday reporting on Verizon Media, 2019; treat as a widely-cited estimate, not a peer-reviewed figure). Regardless of the precise number, autoplay-muted is the default state of TikTok, Instagram Reels, YouTube Shorts, Facebook, and LinkedIn feeds — captions are the only way to convey dialogue without a viewer action.

SEO and content indexing. Search engines, LLM crawlers, and platform recommender systems index caption tracks. A captioned video is a substantively larger surface for topical retrieval than an uncaptioned one. YouTube's search and recommendation systems have used caption text since the mid-2010s.

Non-native speakers and noisy environments. Captions materially improve comprehension for viewers watching in a second language and for anyone in a shared or noisy environment where audio is impractical.

SDH (Subtitles for the Deaf and Hard-of-Hearing) is content that behaves like closed captions but is delivered through a subtitle track. It includes speaker labels and non-speech audio cues but uses subtitle file formats (SRT, WebVTT, TTML) rather than the CEA-608/708 broadcast caption stream. It emerged as streaming platforms consolidated their internal delivery pipelines on TTML/IMSC and needed a same-language track that included accessibility metadata without going through broadcast caption infrastructure.

Use SDH when: you are shipping video to a streaming platform (Netflix, Prime, Apple TV+, Disney+) that expects TTML/IMSC delivery; you want a single format that works across web, mobile, and TV apps; you are producing multilingual content and want a same-language accessibility track alongside translated subtitle tracks.

Compliance treatment. For ADA, WCAG, Section 508, and EAA compliance purposes, SDH and CC are treated equivalently: both satisfy the “captions for synchronized media” requirement as long as the content includes speaker IDs and non-speech audio cues.