GlossCap

Regulatory reference · 21st CVAA · 47 USC § 613 · 47 CFR § 79.4 · 47 CFR § 79.103

CVAA captions: 47 USC § 613, 47 CFR § 79.4, IP-distributed video accessibility

The 21st Century Communications and Video Accessibility Act (CVAA) is the FCC-administered captioning regime for video distributed over the internet that previously aired on US television with captions. It is the most-misunderstood-by-non-broadcast-customers of the federal captioning regimes, because it sits in 47 USC alongside the rest of the Communications Act and is enforced by the FCC rather than the Department of Justice or the Office for Civil Rights — meaning the enforcement pattern and penalty structure differ markedly from the ADA, Section 504, and Section 1557 regimes most L&D and accessibility teams have been preparing for. CVAA covers a narrower content category than ADA Title II — IP-distributed previously-televised programming and certain user-generated-content scenarios — but the substantive caption-quality standard at 47 CFR § 79.4(c) is operationally similar to WCAG 2.1 SC 1.2.2: accuracy, synchrony, completeness, placement. Companies that use broadcast-derived clips in training video — news segments, sports clips, documentary excerpts, syndicated educational content — need to understand which clips fall in scope, how the substantive bar applies, and where CVAA stacks with ADA / Section 504 / Section 1557.

TL;DR

The CVAA's IP-captioning rules are at 47 USC § 613 (statute) and 47 CFR § 79.4 + § 79.103 (FCC regulations). Substantive bar lives at 47 CFR § 79.4(c)(1)–(4): accuracy, synchrony, completeness, placement. Scope is video that aired on US TV with captions and is then redistributed over IP — covered programming. Out of scope: born-IP content (originally produced for the web with no TV airing), most internal training video, most user-generated content (with carve-outs), most foreign-produced content. Enforcement: FCC, not DOJ or OCR. Compliance is the responsibility of the video programming distributor (VPD), video programming provider (VPP), and the video programming owner (VPO) — most enterprises are not in any of these categories. Where CVAA matters for L&D / training: when a company embeds a previously-televised clip into training video, the embed inherits the caption obligation; when a company is itself a VPD/VPP/VPO (some media-and-entertainment, sports-rights-holder, broadcast-news, and educational-content companies are); when an institution wants to map its caption-quality bar onto a single substantive standard that satisfies CVAA, ADA Title II, Section 504, Section 508, and EAA at once.

The statute and the FCC rules

The CVAA was enacted as Public Law 111-260 on 2010-10-08, signed into law as the consolidated update of the 1990s-era accessibility provisions of the Communications Act for the broadband era. Title I of the CVAA addresses telecommunications accessibility; Title II addresses video programming captioning, video description, and emergency-information accessibility. The Title II video provisions land in the Communications Act at 47 USC § 613.

The FCC's implementing rules for IP captioning are at 47 CFR § 79.4. The substantive caption-quality standards were adopted in the FCC's 2014 Closed Captioning Quality Order (FCC 14-12, released 2014-02-24). The user-generated-content and consumer-generated-media provisions are at 47 CFR § 79.103. The TV closed-captioning rules (which predate CVAA and which CVAA references) are at 47 CFR § 79.1.

Key dates in the IP-captioning ramp:

The FCC continues to issue clarifying orders, but the rule architecture has been stable since 2017.

Scope: what CVAA covers and what it doesn't

The single most important question for any company evaluating CVAA exposure is the scope question. The rule covers "video programming previously shown on US television with captions, redistributed over IP." The key elements:

The categories that are not in CVAA scope:

Most enterprise L&D content is born-IP and outside CVAA scope. The substantive accessibility bar for that content is set by ADA Title II / Section 504 / Section 508 / EAA / Section 1557 and the institutional caption policy — not by CVAA.

The substantive caption-quality standard at 47 CFR § 79.4(c)

The four-prong substantive standard the FCC adopted in 2014 — and which applies to IP-distributed video in CVAA scope — is operationally similar to the ADA Title II / Section 508 / EAA substantive bar. The four prongs:

The four prongs were ported wholesale into the FCC's 2014 Order from the National Association of Deaf and Hard of Hearing organisations' substantive standard. The FCC enforces them through complaint-based investigation; substantial complaints can trigger forfeiture orders.

Who is liable: VPD, VPP, VPO

The CVAA's compliance liability splits across three statutory roles defined at 47 USC § 613 and 47 CFR § 79.4(a)(2)–(4):

For most enterprises, the question is whether they are any of these three. Most L&D, marketing, customer-academy, and training-video operations are not — the company isn't distributing TV-aired-with-captions programming. The companies that are:

For everyone else, the CVAA is not the operative regime. The operative regime is ADA Title II / Section 504 / Section 508 / EAA / Section 1557 / state-equivalents, applying to the company's web content broadly.

Where CVAA touches L&D and training video

Even though most L&D content is born-IP and outside CVAA scope, several scenarios create CVAA-relevant questions for training-video operations:

The practical L&D / training-video posture: caption everything substantively to the SC 1.2.2 / § 79.4(c) standard regardless of which regime applies, because (a) the regimes overlap on the substantive bar, (b) the unified policy is operationally simpler than slicing content by regime, and (c) the substantive bar is what produces a usable caption regardless of regulatory framing.

How CVAA stacks with the rest of the federal captioning regime

For an institution that produces or hosts a mix of content, the captioning regimes stack:

The substantive caption-quality bar is similar across the regimes. The differences are in scope, enforcement body, and penalty structure. A unified institutional caption-quality policy that meets SC 1.2.2 substantive accuracy will clear most regimes' substantive bars.

FCC enforcement pattern and penalty structure

FCC enforcement of CVAA captioning is complaint-driven. The pattern when a complaint lands on a VPD/VPP/VPO:

  1. Consumer files complaint. Complainant identifies the programming, the distribution platform, and the caption failure observed. The complaint can be filed informally (FCC consumer complaint portal) or formally.
  2. VPD response window. The VPD has a 30-day response window (47 CFR § 79.4(d)) to investigate and either correct the issue or document why it doesn't apply.
  3. FCC Consumer and Governmental Affairs Bureau review. If the consumer is unsatisfied with the VPD response, the complaint advances within the FCC.
  4. Possible Enforcement Bureau referral. Substantial pattern complaints or wilful violations are referred to the Enforcement Bureau.
  5. Forfeiture order or consent decree. The Enforcement Bureau can issue a Notice of Apparent Liability, leading to a forfeiture order or a negotiated consent decree.

The penalty structure differs from the ADA / Section 504 / Section 1557 regimes. CVAA forfeiture orders carry monetary penalties (the FCC's standard penalty schedules) plus required corrective-action plans plus reporting obligations. The DOJ's settlement authority under ADA Title II / Title III is broader; the OCR's enforcement under Section 504 / Section 1557 includes loss of federal funds. The CVAA's penalty structure is monetary and corrective without the funding-loss leverage.

The user-generated-content carve-out at 47 CFR § 79.103

The FCC's user-generated-content provision at 47 CFR § 79.103 provides a partial carve-out for video uploaded by consumers to video hosting platforms. The rule recognises that consumer-uploaded video typically does not include captions and that imposing captioning on consumer uploads would impose substantial cost on the upload platform. The carve-out is narrower than it sometimes appears:

For institutions that allow employees or students to upload video to internal video hosts (an internal Vimeo, an internal Stream, an internal Wistia channel, an institutional YouTube channel), the user-generated-content carve-out provides limited cover. Most of that content is still in scope under ADA Title II / Section 504 / Section 1557 / EAA when the institution distributes it; CVAA's user-generated carve-out doesn't lift those obligations.

Where glossary-biased captioning meets the § 79.4(c) bar

The § 79.4(c)(1) accuracy prong is the prong that generic auto-captioning fails most often. The proper-noun mangling that fails SC 1.2.2 fails § 79.4(c)(1) the same way: the regulatory citations a healthcare news segment is reporting on, the named characters in a documentary excerpt, the technical product terms in a sports analyst's commentary, the political proper nouns in a news interview — these are the words that distinguish substantively accurate captions from substantively inaccurate ones, and they are the words generic STT has the least training data for.

Glossary-biased captioning lifts the accuracy prong specifically. The institution builds the glossary once — programme names, recurring proper nouns, regulatory citations, named characters in long-running content — and the per-asset accuracy compounds across the catalogue. For a VPD with a sustained back-catalogue captioning workload, the cost-per-hour collapses substantially relative to the hand-corrected auto-caption pipeline or the human-vendor pipeline at $1.25-$3.00 per minute. Vendor pricing breakdown.

See pricing

FAQ — CVAA captions

Does CVAA apply to my company's training video?

Almost certainly not, if your training video is born-IP (produced for the web with no TV airing). CVAA covers video previously shown on US television with captions and then redistributed over IP. Most L&D / training video is outside that scope. The operative regime for your training video is ADA Title II / Section 504 / Section 508 / EAA / Section 1557 / institutional accommodation policy — not CVAA.

Does CVAA apply if I embed a TV-news clip in my training video?

The clip itself is in CVAA scope when it was previously shown on US TV with captions and is being redistributed over IP. The captioning obligation flows through the licensor; the institution's responsibility is to verify the clip carries captions when sourced and to keep them in place. The institutional training video as a whole is in scope of the regime that applies to its distribution channel (ADA Title II for a public university, Section 1557 for a covered healthcare entity, EAA for an EU operation).

What is the substantive caption-quality standard under CVAA?

47 CFR § 79.4(c) — accuracy, synchronicity, completeness, placement. The four-prong test was adopted by the FCC's 2014 Closed Captioning Quality Order and is operationally similar to SC 1.2.2 substantive accuracy. The placement prong is unique to CVAA (SC 1.2.2 doesn't address visual placement directly).

How does CVAA compare to ADA Title II?

Different scope, different enforcement body, different penalty structure. CVAA covers IP-distributed previously-televised programming and is enforced by the FCC. ADA Title II covers state and local-government public-entity web content broadly and is enforced by the DOJ. The substantive caption-quality standards align operationally; the procurement-and-distribution context differs.

Are streaming services like Netflix and Disney+ subject to CVAA?

Yes — they are VPDs distributing video programming over IP. CVAA's IP-captioning rules apply to their distribution of in-scope programming. The streaming services' obligations under CVAA are an active area of FCC enforcement and consumer complaints.

What about live and near-live programming?

Live programming aired on TV with captions and redistributed over IP is in scope under the 2013-03-30 deadline. The accuracy and synchronicity prongs of § 79.4(c) apply differently to live content (FCC guidance recognises the inherent difficulty of live captioning), but the obligations stand. Live captioning typically uses a CART captioner or a real-time STT system; the substantive accuracy expectations are calibrated to live conditions.

What about archival broadcast content that pre-dates the 2012 effective date?

Archival broadcast programming that aired with captions and is now redistributed over IP picks up CVAA obligations under the post-2012 deadlines. The captioning obligation applies prospectively to redistribution after the effective date, regardless of when the original TV airing occurred.

Does CVAA apply to user-generated content?

Genuinely user-generated content is outside CVAA scope under the § 79.103 carve-out. User uploads that incorporate substantial in-scope programming (mash-ups containing significant portions of TV-aired-with-captions content) may pull the user-generated portion partly into scope. Most YouTube / Vimeo / TikTok / Loom uploads of original consumer-generated material are outside CVAA scope; the institutional captioning bar for that content lives in Section 504 / ADA Title II / EAA when the institution embeds or hosts it.

Where does CVAA fit alongside the FCC's TV closed-captioning rules?

The TV closed-captioning rules at 47 CFR § 79.1 set the captioning obligation for the original broadcast — the rules that originally drove the captions onto the TV airing. CVAA / 47 CFR § 79.4 then carries that captioning obligation through to the IP redistribution. The two rule-sets work together: § 79.1 captions the broadcast, § 79.4 carries the captions through to the streaming or on-demand distribution.

Further reading