# WCAG 3 Conformance vs. Bronze, Silver, Gold: How to Claim a Tier
> WCAG 3 conformance vs. Bronze, Silver, Gold reporting tiers in the September 2026 draft: what differs, the four steps to claim one, and why no one can yet.
**Published:** 2026-10-18 | **Updated:** 2026-10-18
---
When "pass" and "grade" show up in the same document, it's tempting to put them on one line: the passing mark is the lowest grade, and the grades climb from there. Most exams work that way. WCAG 3's conformance and Bronze, Silver, and Gold are easy to read the same way.
On October 16, I gave a talk, [What the WCAG 3.0 Draft Asks of Evaluation Methods]({{< relref "/talks/2026-nia-wcag3" >}}), at a meeting of the Digital Accessibility Standardization Forum's service subcommittee. In the Q&A, someone asked: **"What's the difference between meeting conformance and Bronze, Silver, Gold? And what do you have to do to get them?"** I couldn't answer it well with only the materials I'd prepared, and I've regretted that since. So I went back to the WCAG 3 Working Draft published on September 10, 2026, with just that one question in mind. WCAG 3 is the standard W3C is building to succeed WCAG 2.2, and a Working Draft is one that isn't final yet.
{{< eli5 >}}
Think about graduating from university. Skip one required course and no number of electives will get you the diploma. Graduating with honors is a distinction added only for graduates who clear a higher bar. In the WCAG 3 September draft, conformance is the diploma, and Bronze, Silver, and Gold are the honors. The twist: the WCAG 3 draft names no registrar for the honors; it only describes claiming them once you meet the bar, and the grading rubric is still blank. After reading this post, you'll be able to explain conformance and the tiers separately, what it takes to claim a tier, and what you can prepare right now.
{{< /eli5 >}}
{{< img src="images/contents/honor-cords-graduation-cap.jpg" alt="A graduation cap beside yellow honor cords and a diploma - honors go only to people who have graduated, and Bronze, Silver, and Gold only count once you conform" caption="Photo: Susan Weber / Unsplash" >}}
## The 30-Second Answer: Conformance Is the Passing Line, Bronze, Silver, and Gold Are Reporting on Top
Answer first. If I could go back to that Q&A, this is what I'd say.
> Conformance is the verdict on whether you met every core requirement; Bronze, Silver, and Gold report what you did beyond that. To claim Bronze, you first have to conform. On top of that you add supplemental requirements, plus assertions stating that you ran procedures such as usability testing, in numbers that haven't been set yet. Silver takes more, and Gold also needs organization assertions. The draft doesn't define a body or review process that awards tiers; it describes the tier being claimed by whoever meets it. But the numbers are blank, and so are some of the threshold values that determine conformance, so nobody can claim a tier from the September draft. For now, the preparation is to meet WCAG 2.2, and to run those procedures and keep records of them.
Put briefly: **conformance asks "does it meet the standard?", while Bronze, Silver, and Gold report "how much more did you do after meeting it?"** Below, I'll check each part of that answer against the draft's own wording.
## Core, Supplemental, Assertion, Reporting Tier: Getting the Terms Straight First
If WCAG 3 is new to you, a few terms will look unfamiliar. This post rests on the six below. For more detail, see [What Changed in the WCAG 3 September 2026 Draft]({{< relref "/posts/wcag-3-draft-september-2026" >}}).
- Conformance means the content meets the standard's requirements. WCAG 2.2 splits its success criteria into three levels, A, AA, and AAA, and most teams aim for AA.
- Core requirements are the ones you must meet to conform. The September draft has 106.
- Supplemental requirements aren't needed for conformance and sit on top of the core set. There are 74.
- Assertions are documented statements that "we carried out this procedure." There are 35.
- Tags rate each core requirement by the worst harm that could result from missing it: Physical harm, Risk, Barrier, or Friction. Assertions get content or organization tags. In the September draft, none of these have been assigned to individual items yet ([what the four tags mean]({{< relref "/posts/wcag-3-draft-september-2026#the-four-tags-for-core-requirements--how-bad-is-it-if-you-miss-one" >}})).
- Reporting tiers are a six-step scale for saying how far an evaluation result gets. There are two steps below conformance, split by tag (tier 1, Avoid physical harm, and tier 2, Foundational access), tier 3 is conformance, and the three above it are Bronze, Silver, and Gold.
> The talk itself was in Korean and used slightly different Korean terms for several of these; this post follows the terminology used across this series. When the talk described the March draft's Bronze/Silver/Gold as conformance levels "like A/AA/AAA", that meant there were three levels, not that Bronze equals A.
## Conformance and Bronze, Silver, Gold Answer Different Questions
The main reason the two get confused is that they appear in the same table. In the reporting tier table in section 4.1.4 of the September draft, tier 3 is conformance and tiers 4, 5, and 6 are Bronze, Silver, and Gold. Put them on one staircase and they naturally look like one line. But follow the source text and you'll see they live in different chapters of the document: the conformance rule is in the chapter you must follow, and the tiers are in an advisory chapter.
Conformance is in chapter 3 (Conformance), and the rule is a single sentence.
> For a specified conformance scope to conform to WCAG 3, all core requirements MUST be satisfied.
That is, every core requirement in your chosen conformance scope must be met (MUST) for you to conform to WCAG 3. An editor's note in section 3.2 (a memo recording points the working group is still discussing) says directly what changed in this version.
> This update to WCAG 3 moves the concept of leveling from conformance to reporting. To keep this clear, we refer to "reporting tiers" to distinguish them from previous "conformance levels".
Leveling moved out of conformance and into reporting, and the new name "reporting tiers" keeps it from being mixed up with the old conformance levels. That's why Bronze, Silver, and Gold live in chapter 4 (Reporting), which introduces itself like this.
> This section presents reporting tiers, not conformance tiers. The third reporting tier equates with conformance to WCAG 3.
Chapter 4 covers reporting tiers, not conformance tiers, and its third step is WCAG 3 conformance. The header of chapter 4 also carries a label saying the whole chapter "provides advice only", meaning it's non-normative. Normative parts are what you must follow to say you followed the standard. Non-normative parts are explanatory, and skipping them doesn't mean you've broken the standard.
{{< img src="images/contents/bsg-two-axes-en.png" alt="A diagram with the verdict on the horizontal axis (not conformant, conformant) and reporting tiers on the vertical axis - on the not-conformant side there are only tier 1, Avoid physical harm (core requirements tagged Physical harm and Risk), and tier 2, Foundational access (adds Barrier), with no tiers above them; on the conformant side, Bronze, Silver, and Gold stack above tier 3, conformance (all core requirements), and the number to add for each is TBD (to be determined)" caption="The verdict has two outcomes, and Bronze, Silver, and Gold only stack up inside the conformant one." >}}
Side by side, it looks like this.
| Aspect | Conformance | Bronze, Silver, Gold |
|---|---|---|
| What it asks | Does this scope conform to WCAG 3? | How much more did you do after conforming? |
| Where it lives | Chapter 3, Conformance (3.2) | Chapter 4, Reporting (4.1.4) |
| Normative? | Normative (MUST) | Non-normative (advice only) |
| Section status | Developing | Exploratory |
| Result | Conformant or not conformant, nothing in between | How far up tiers 4, 5, and 6 you've come |
| What it takes | All core requirements in scope | Conformance + supplemental requirements + assertions |
| Counts | All of the 106 core requirements that apply to the scope | [TBD] (To Be Determined) for each tier |
{caption="Conformance vs. Bronze, Silver, Gold (September draft)"}
Section status is a maturity label the draft puts on each section. Developing means there's rough agreement on what's needed, but the big issues aren't all settled and the details are still being polished. Exploratory means the direction is still being explored, so details and definitions may be missing. The conformance rule can still change, and the tiers can shift even more.
## What It Takes to Claim Bronze, Silver, or Gold: Four Steps
Now for the second half of the question. Broken down to make it easier to follow, the road to a tier has four steps: **conform, meet more supplemental requirements, make assertions, and state it.** That breakdown is mine, not a procedure the draft defines; it's a reading aid for the 4.1.4 table and the top of chapter 4, and steps two and three have no set order.
In the table in 4.1.4, the three boxes above conformance stack up like this. Bronze adds supplemental requirements and content assertions to conformance, a "[TBD number of]" of them. Silver adds a "[TBD greater number of]". Gold adds organization assertions on top of that. On all three rows, the count is blank (TBD).
{{< img src="images/contents/bsg-four-steps-en.png" alt="A left-to-right flowchart of the four steps to claim Bronze, Silver, or Gold - 1 Conform (all core requirements; some threshold values still undecided), 2 Meet supplemental requirements, 3 Make assertions (do the work, fix issues where needed, state it with the required information; tags not yet assigned), 4 Claim it (claimed by whoever meets it; the draft has no review process). The count for steps 2 and 3 is a single slot and how to count is undecided, so nobody can claim a tier yet" caption="The counts for steps 2 and 3 are still blank." >}}
### Step 1: Conform, Meeting Every Core Requirement
The starting point is conformance. Right above the table, section 4.1.4 says this.
> Each tier is cumulative, building on and including all requirements from the tiers before it.
Tiers stack, and each one includes every requirement from the tiers below. Miss even one core requirement and no number of supplemental requirements gets you to Bronze, because tier 4 contains tier 3.
Conformance comes with two more conditions.
One is scope. Conformance is judged only inside the scope you define. It can be the whole site, or narrowed to specific pages or a process such as sign-up, and once you've set a scope, everything testable inside it has to be included (MUST, 3.2.2). An editor's note in the same section says narrower scopes such as paths or components are still being defined. Narrow the scope, and both conformance and tiers become statements about what's inside that boundary.
The other is accessibility support. Only methods that work in real users' browsers and assistive technologies (screen readers, for example) count toward conformance. A conformance claim (a different document from a public accessibility statement) has to say which combinations of browser and assistive technology (the accessibility support set) were used to check (3.2.1.1). I covered this concept in [WCAG 3 Accessibility-Support Sets: Korea's Blank]({{< relref "/posts/wcag-3-accessibility-support-sets" >}}), but that post is based on the March draft, where the same provision is numbered 3.2.2.
### Step 2: Meet More Supplemental Requirements
Section 3.1.1 sums up what supplemental requirements are in three kinds.
> Supplemental requirements — requirements that build on the core set, not required for conformance. These include: requirements that improve the accessibility of content beyond core requirements, a stricter form of a comparable core requirement, and requirements that are disproportionately burdensome to implement.
They aren't needed for conformance, and they come in three kinds: requirements that raise accessibility beyond the core set, stricter versions of comparable core requirements, and requirements that are very burdensome to implement.
The second kind, the "stricter form", is easy to see in the pair of flashing requirements. When a screen flashes rapidly it can trigger photosensitive seizures, so the draft sets thresholds.
| Aspect | No flashing over threshold | No flashing over threshold (no exceptions) |
|---|---|---|
| Kind | Core requirement | Supplemental requirement |
| Requirement | Flashes stay below the general flash and red flash thresholds (including a condition that treats small-area flashes as passing) | Below the thresholds, with no minimum-area condition |
| Exception | When the flashing is essential | None |
So a stricter version of the same topic lives separately on the supplemental side. It's similar to how 2.3.1 Three Flashes or Below Threshold (A) and 2.3.2 Three Flashes (AAA) relate in WCAG 2.2.
How many of the 74, and which ones, make Bronze? **Nobody has decided.** The count slot in the 4.1.4 table is blank, and it doesn't say which supplemental requirements are counted or whether you get to choose. The slot also lists supplemental requirements and assertions together as one number, so it's not yet clear whether you fill them separately or count them together.
### Step 3: Assertions, Where You Actually Do the Work and State It with the Required Information
Assertions cover things you can't learn by inspecting a screen: whether you ran usability testing with users, whether you reviewed your text in plain language. Section 4.1.3 splits assertions with two tags. Content assertions are procedures that check content, such as plain language review, usability testing, assistive technology testing, and style guides, and they're needed for Bronze and Silver. Organization assertions are procedures that help an organization produce accessible content, and they're needed for Gold. The source gives staff accessibility training and a public accessibility statement as examples of organization assertions, but **none of the 35 assertions in the September draft body has either name.** The definitions of the two tags are in [the assertion tags section of the September draft post]({{< relref "/posts/wcag-3-draft-september-2026#two-assertion-tags--content-and-organization" >}}).
{{< img src="images/contents/checklist-tablet-pen.jpg" alt="A hand marking a checklist on a tablet with a pen - an assertion is a statement, with the required information, that you carried out a procedure or practice to improve accessibility" caption="Photo: Jakub Żerdzicki / Unsplash" >}}
To see what a single assertion actually looks like, look at "Captions usability testing" in the September draft body.
> [Title, role, or organization] asserts that: We have conducted usability testing with users who need captions, and changes were made to fix or mitigate the issues found.
The format states that the named person, role, or organization ran usability testing with users who need captions and fixed or reduced the problems they found. Below it, the draft separates what to publish from what to keep internally.
| Category | Items |
|---|---|
| Information to publish | The person or organization making the statement (if different from the one declaring conformance), the date of the usability testing, the date of the statement (if different from the conformance claim date) |
| Internal records recommended (non-normative) | Scope, each participant's disability type, number of participants per disability type, test date, issues found and solutions applied |
{caption="Public information and internal records for Captions usability testing"}
What evaluators look at is in section 6.5 of the [Explainer for WCAG 3.0](https://www.w3.org/TR/2026/DNOTE-wcag-3.0-explainer-20260910/) (the Explainer from here on, a non-normative document describing WCAG 3's structure and background).
> The results of the process stated in an assertion cannot be tested. Instead, the quality of an assertion statement can be tested based on how well the assertion meets the documentation requirements for assertions (see Documenting Assertions).
The results of the process can't be tested, so what gets tested is how well the statement meets the documentation requirements. The same section adds that supporting records aren't tested for conformance either, and Explainer 5.3.1 says assertions cover things that "cannot be verified by independent testers". What evaluation checks is whether the statement meets the documentation requirements; it doesn't ask anyone to verify the results of the procedure or the supporting records.
That doesn't mean you can run the captions usability test and stop, though. This assertion's statement contains "changes were made to fix or mitigate the issues found". If you didn't fix anything, you can't write that sentence. So for the captions usability testing assertion, there are three things to do: **actually run the test, act on what you find, and state it with the required information.** Not every assertion looks like this. Quite a few, like the style guide and review policy ones, state that "we have this policy and process", so what you have to do depends on each statement.
The 35 assertions in the September draft don't carry content or organization tags yet. Author training shows up only inside a few assertions, such as Clear language review, as a condition ("if you train authors") or a recommended internal record. Judging by the examples in 4.1.3, almost none of the ingredients for Gold exist yet.
> The post that first covered assertions, [Assertions: A New Unit for Accessibility Evaluation]({{< relref "/posts/wcag-3-assertions" >}}), was written against the March draft. Its concept explanation still reads fine, but for the counts (28 in March, 35 in September), the role assertions play in tiers, and the description of assertions being judged True/False, please use this post (Explainer 6.5) as the reference.
### Step 4: Claiming It, and the Draft Has No Body That Reviews Tiers
The last step is stating it. The draft names no body that reviews and awards Bronze, Silver, or Gold. An editor's note at the top of chapter 4 gives this example.
> For example, assertions included in tiers beyond conformance could be completed in order to claim a Silver achievement in WCAG.
Complete the assertions in the tiers beyond conformance and you can claim (the note's word) a WCAG Silver achievement. Section 7 of the Explainer points the same way.
> Meeting a certain number or percentage of the supplemental requirements and assertions enables you to claim that tier.
Meet a set number or percentage and you can claim the tier. A conformance claim isn't mandatory, and you can conform without making one (3.2.2). Explainer 5.3.2 says assertions used for conformance must be written in the conformance claim, and when they're used elsewhere they can be published somewhere like the organization's website. The draft defines no body or review process for tiers; it only describes the party that meets the criteria stating the scope and evidence and claiming the tier. It doesn't rule out third-party evaluation either, so who checks these claims in practice will depend on future policy and practice. How a tier would be displayed isn't defined either.
But "no body awards tiers" doesn't mean "nobody will require them". The note at the top of chapter 4 includes an example where the working group could recommend that regulators require Silver of organizations above a certain size. An editor's note in 4.1.4 says the same thing.
> As a standard, WCAG declares conformance when the core requirements are met. Policymakers have more flexibility in requiring additional tiers in certain situations in order to comply with regulation.
WCAG as a standard only declares conformance when the core requirements are met. Whether to require a tier above that is up to policymakers. If a law or a procurement condition says "Silver or above", the tier becomes mandatory in that setting.
## The Current Draft Doesn't Let You Claim Bronze Yet: What You Can Prepare Instead
{{< img src="images/contents/empty-podium-stage.jpg" alt="An empty 1-2-3 podium under stage lights - the podium is built, but the criteria to engrave on the medals are still blank" caption="Photo: Joshua Golde / Unsplash" >}}
The medals have names, but the criteria to engrave on them are still blank. Somebody built the podium before writing the rules of the race. There are three reasons you can't claim Bronze from the September draft.
1. **There are no counts.** The Bronze, Silver, and Gold rows in the 4.1.4 table are all TBD. If you don't know how many to meet, you can't say you've met them.
2. **Assertions have no tags.** Bronze and Silver count content assertions and Gold counts organization assertions, but which of the 35 is which hasn't been assigned. The Physical harm, Risk, Barrier, and Friction tags on core requirements are blank too, though those matter for reporting tiers 1 and 2, below conformance.
3. **Even the conformance verdict can't be reached.** Text contrast reads "@@[contrast measure to be determined]", and letter shapes and paragraph display read "@@ [values to be determined]", so some core requirements have blank threshold slots. Today you can rule non-conformance out, but you can't confirm conformance. For the accessibility support set a conformance claim must name, the default set isn't defined yet either, and there's no guidance yet on how to build an alternative set for a language like Korean (editor's note, 3.2.1.1).
Those three are blanks that keep a verdict from being reached. One more thing worth knowing is the status of the framework itself: chapter 4 on reporting tiers is non-normative and Exploratory, so it may change even after the counts are filled in. The Explainer also carries an alternative that shows progress as a score instead of the steps below conformance.
So what can you do now?
**Meet WCAG 2.2.** The WAI [WCAG 3 introduction](https://www.w3.org/WAI/standards-guidelines/wcag/wcag3-intro/) opens with this line.
> The best way to prepare for WCAG 3 in the future, is to meet WCAG 2.2 success criteria now.
Section 1.2 of the September draft also expects that content meeting WCAG 2.2 A and AA will mostly meet 3.0 conformance too, with the caveat that added testing and different scoring mean some extra work. Since 3.0 conformance is the floor for every tier, keeping to 2.2 AA is also the first step toward Bronze.
**Run the procedures that would become assertions, and keep records.** If you wait until the counts are set to start, you'll be late. Procedures like usability testing, where you have to recruit participants and line up schedules, don't happen overnight. If you model your records on the items in the draft, you'll know what to write down.
| Procedure | Records to keep |
|---|---|
| Usability testing | Topic, scope, each participant's disability type, number of participants per disability type, test date, issues found and solutions applied |
| Plain language review | A copy of the review policy and the date it was put into effect, dates of author training and the number or share of authors who completed it, a copy of the style guide |
| Style guide | A copy of the guide, the policy that requires following it, the date the guide was published, a summary of the user needs it reflects, issues found and solutions applied |
| Assistive technology testing | Assistive technologies and browsers tested (product and version), dates, issues found and actions taken |
| Staff training, accessibility statement | Training dates and number of people who completed it, where the statement is published and when it was last updated |
{caption="Records to keep for each procedure, in preparation for assertions"}
> The first three rows come from the items in the September draft's Captions usability testing, Clear language review, and Captions style guide (some are public items). The last two rows are examples that 4.1.3 mentions, but no assertion with those names exists among the 35, so I filled them in using the same pattern.
Usability testing assertions exist per topic, such as captions, audio description, key information, and help, and the topic decides who the participants are. Records from your existing user testing can be carried over only to topics whose participant types match.
**Check the supplemental requirements against your site.** Mark which of the 74 you already meet, and when the counts are set you can see right away how many are left. Several in the text display family still have blank threshold values even on the supplemental side, so leave those cells empty.
**Record severity in two layers in your self-assessment report.** In 3.0, tags are assigned per requirement, based on the worst harm that could result from missing it. If you note, for each checkpoint, whether the worst case is physical harm, risk, barrier, or friction, and separately note the actual impact of each error, it'll be easy to map your records over when tags are assigned. The editor's note in section 4.1 also says the same requirement can differ in severity depending on where the error occurs and in what context.
Still, it's too early to write 3.0 conformance, non-conformance, or tiers in a report as a verdict. Some core threshold values and all the tier counts are blank. How to line up your evaluation records with 3.0's unit of judgment (which browser and assistive technology you checked with, and a scope defined by a process such as sign-up) is covered in question 24 of the practical reference materials on [the talk page]({{< relref "/talks/2026-nia-wcag3" >}}).
## Bronze Under the March vs. September Drafts: Walking Through the Same Sites
The clearest way to see what the September draft changed is to judge the same two sites under both drafts. The two sites below are hypothetical, and I'm assuming every blank value in the September draft has been filled in, just following the rules.
- Site A meets every core requirement in scope. It meets no supplemental requirements and has no documented procedures (assertions).
- Site B misses one core requirement. In return, it meets several supplemental requirements and ran usability testing with users with disabilities and fixed the problems it found.
Section 3.2.1 of the March draft set the lowest conformance level as Bronze and put it like this (the model is covered in detail, against the March draft, in [WCAG 3.0 Conformance Model: Moving Beyond A/AA/AAA]({{< relref "/posts/wcag-3-scoring-conformance" >}})).
> For Bronze conformance (the minimum level of conformance), all of the pages/views and processes within the conformance scope MUST satisfy all the core requirements and some portion (to be determined) of the supplemental requirements and assertions within each functional performance statement.
In the March draft, Bronze conformance took all core requirements plus some portion of the supplemental requirements and assertions (the share to be determined) within each functional performance statement. A functional performance statement is a sentence describing a different functional limitation, such as vision, hearing, or cognition. The September draft's 4.1.4 table has no such condition. Apply this sentence to the two sites, and as long as you read "some portion" as more than zero, the results split like this.
{{< img src="images/contents/bsg-march-vs-september-en.png" alt="Side by side, the results of judging Sites A and B under the March and September drafts - the content is the same as the table right below" caption="By the rules, a site that meets every core requirement conforms under the September draft." >}}
| Aspect | March draft (March 3, 2026) | September draft (September 10, 2026) |
|---|---|---|
| Site A | Not conformant. All core requirements met, but fell short of Bronze's "some portion" condition | Conformant (tier 3). Nothing to report for Bronze, Silver, or Gold |
| Site B | Not conformant. One core requirement missed, so no Bronze | Not conformant. No Bronze either. Reporting tiers 1 and 2 become possible once tags are assigned |
{caption="Judging Sites A and B under the March and September drafts (hypothetical)"}
What changed is Site A. Under the March draft, meeting all core requirements wasn't enough if supplemental requirements and assertions fell short. Under the September draft, the same site conforms, and Bronze, Silver, and Gold become a separate matter of reporting. Site B doesn't conform under either draft. Explainer 5.3.1 says it outright: "Assertions cannot be used as alternatives to core requirements." Lots of supplemental requirements and usability testing still don't fill the hole left by one missed core requirement.
> For a hypothetical showing how far a site that doesn't conform yet, like Site B, has come on tier 1 (Avoid physical harm) and tier 2 (Foundational access), see [the tier 1 and 2 calculation example in the September draft post]({{< relref "/posts/wcag-3-draft-september-2026#walking-through-it--what-tier-would-your-site-land-on-a-hypothetical" >}}). This post covers only what's above conformance.
## Two Lingering Worries About Assertions: False Statements and the Burden on Small Organizations
Two questions kept nagging at me while I was putting this together. If evaluation only checks whether a statement meets the documentation requirements, couldn't someone write a false one and get away with it? And can a small or mid-sized site or company really afford an assertion that requires usability testing? When I went back through the source text and the working group's repository, both turned out to be well-founded worries, and both are questions with no settled answer yet.
### If Nobody Verifies Them, Couldn't Someone Just Lie?
As we saw earlier, Explainer 6.5 says the results of the stated procedure can't be tested, and 5.3.1 says assertions cover things independent testers can't verify. In other words, this evaluation alone can't guarantee that the procedure was actually carried out or that its results hold up. And nowhere in the September draft or the Explainer does it say how false statements would be handled.
What the source does have is this. You have to publish the person or organization making the statement and the date, so a public record shows who said what and when. Explainer 6.5 says an organization may add its own record-keeping requirements to fit its process, and an editor's note at the end of 5.3.2 puts "What optional supporting documentation should organizations provide with an assertion?" up as a question for comment. How much supporting evidence to ask for hasn't been decided.
Similar worries are on the working group's repository. [Issue #473](https://github.com/w3c/wcag3/issues/473) asks, about the usability testing assertion, "What if it was 10 years ago when the tests were conducted?" That's not a lie, but it's a current statement resting on an old test. As of October 17, 2026, the issue is open. [Issue #816](https://github.com/w3c/wcag3/issues/816) goes a step further: "A company can truthfully assert that it performed usability testing with people with disabilities and still release an inaccessible product." A true statement doesn't guarantee the outcome (this issue was written about the March draft model, but the point applies to the September draft as is).
It's worth noting how far the damage can reach, though. In the current draft, assertions don't enter the conformance verdict, so a false assertion can't turn a non-conformant site into a conformant one. What can be inflated is a claim of Bronze or above. I think the problem gets much bigger the moment a policy starts requiring a tier like Silver. On [issue #634](https://github.com/w3c/wcag3/issues/634), a working group participant left a comment (September 24, 2026) saying a task force to handle assertions is just getting started.
### Can Small Sites and Companies Afford This?
The working group is asking this one directly. An editor's note at the end of Explainer 5.3.2 lists it among the questions it wants comments on.
> How can small organizations use assertions without unrealistic burden?
So there's no answer yet on how small organizations could use assertions without an unreasonable burden. The editor's note at the top of chapter 4 in the September draft also gives an example of regulators requiring Silver of "organizations of a certain size", which shows organization size is on the working group's mind.
The burden falls less on the information you publish than on the procedure. In the captions usability testing example earlier, what you publish is three fields, but a usability testing assertion has to be done with the users the topic specifies: users who need captions for captions, or participants including people with cognitive or mental health disabilities for key information and help. For a small organization with no dedicated staff, even recruiting participants isn't easy. Issue #816 also notes that "a small organization can produce an accessible product without possessing the formal governance apparatus of a multinational enterprise", which means how well you've set up procedures and how accessible the product is can come apart.
Still, the starting line is the same for everyone. Assertions aren't needed for conformance, so even a small organization can conform on core requirements alone. The burden grows when you try to report above Bronze. Both worries are open questions with no settled answer, which makes them comments you can submit through the channels below.
## Five Common Misconceptions About WCAG 3 Bronze, Silver, and Gold
Reading the source text, I picked out five places where it's easy to get tangled.
### Is Bronze the Same as WCAG 2.2 Level A?
No. In the September draft, Bronze is a tier you can only report once you conform, so it can't sit in the same slot as A, which is one of the conformance levels. In the March draft too, all three tiers were built on all core requirements. If you want a pairing, the closest is 2.2 AA conformance and 3.0 conformance (all core requirements), and the three tiers resemble AAA only in being above the passing line. There are other differences. Assertions, which 2.2 didn't have, come in, and tiers are split by how many extras you've done rather than by meeting a full set of criteria.
### If You Don't Have Bronze, Are You Not Conformant?
That was true through the March draft, but not in the September draft. Meet all core requirements and you conform without Bronze. Conversely, if you don't conform, there's no Bronze.
### Are Tiers 1 and 2 Partial Conformance?
No. Tier 1 (Avoid physical harm) and tier 2 (Foundational access) are steps for reporting where you are on the way to conformance. Conformance is tier 3 only, and a site at tier 1 or 2 doesn't conform yet. Explainer 7 also places these two tiers under "On the way to conformance".
### Is a Reporting Tier the Same as WCAG-EM's Reporting Step?
No, they're different things. Step 5 of [WCAG-EM 2.0]({{< relref "/posts/wcag-em-2-evaluation-methodology" >}}) (W3C's accessibility evaluation methodology), Report the evaluation findings, is the procedure of writing up results after the evaluation is done. WCAG 3's reporting tiers are the scale for the results you can write in that report. One is a step in the evaluation sequence and the other is a way to bin the results. Calling both a "reporting stage" invites confusion, so this series says "reporting tier".
### If the Scoring Alternative Is Adopted, Do Bronze, Silver, and Gold Disappear?
No. The scoring alternative in Explainer 7.1 (Alternative approach - Scoring) weights the tags and shows progress toward conformance as a percentage (for the weights, see [the scoring alternative section of the September draft post]({{< relref "/posts/wcag-3-draft-september-2026#a-scoring-alternative-left-open--in-the-explainer-only-not-the-draft" >}})). Even in that approach, once you conform, it counts supplemental requirements and assertions and "awards" Bronze, Silver, and Gold. The two approaches differ mainly in how they show what's below conformance (the Gold composition differs a bit too), and they agree that three tiers sit above conformance.
## Passing Korea's Web Accessibility Quality Certification Is Not WCAG 3 Conformance
If you work in Korea, this is probably where web accessibility quality certification comes to mind. It's a national scheme run by NIA, the National Information Society Agency, through designated certification bodies, which review a site and award a certification mark, and it's based on the Korean Web Content Accessibility Guidelines 2.2 (KWCAG 2.2, KS X OT0003). **Passing certification and WCAG 3 conformance reach their verdicts in different ways.** NIA's "Standard Review Guidelines for Web Accessibility Quality Certification Bodies" (December 2024, Korean only) counts a site as passing only if it clears both expert review and user review. I've laid out the review procedure and the re-review conditions in [WA Certification: Cost, Process, How to Get It]({{< relref "/posts/web-accessibility-certification-guide" >}}).
- In expert review, each of the 33 checkpoints in KWCAG 2.2 has to reach a compliance rate of at least 95% **on its own**. The rate for each checkpoint is "compliant pages ÷ pages reviewed" (content count for the captions checkpoint). The default sample for a new review of a medium or large site is 20 to 25 pages, so in the first round, for checkpoints counted by page, a checkpoint marked down on more than one page fails.
- In user review for a new certification, reviewers with disabilities of three types (blind, low vision, and upper-limb physical or brain lesion disabilities) have to complete each task, as a rule, within 15 minutes (complex tasks can be exempted), with a 100% task success rate for each disability type. If the tasks were completed, it counts as a pass (with recommendations) even when the reviewers record inconvenience notes or recommendations.
WCAG 3 conformance holds only if every core requirement in scope is met, and there's no per-item percentage line to clear. Conversely, certification has something 3.0 conformance doesn't: reviewers with disabilities have to actually complete tasks. In the September draft, usability testing isn't a conformance requirement but an assertion for the tiers above it (whether to make it mandatory is undecided, per the 4.1.3 editor's note). So the same site could come out either way.
There's a naming trap too. The 불편 사항 (literally "inconvenience notes") in Korean user review and the 3.0 tag Friction look alike only because the Korean-language talk rendered Friction as the same word, 불편. They sit in completely different places. In Korean certification, 불편 사항 is a result marker in user review meaning the task was completed but with some inconvenience, while whether a checkpoint was violated is decided separately by expert review. Friction in 3.0 is a tag on requirements, so a core requirement tagged Friction still makes you non-conformant if you miss it. Don't read one as the other just because the names resemble each other.
{{< img src="images/contents/bsg-kr-cert-vs-wcag3-en.png" alt="A side-by-side diagram of Korea's web accessibility quality certification pass criteria and WCAG 3 September draft conformance criteria - the criteria for each side are the same as the list right above, and the bottom line says a certification pass can't be translated into WCAG 3 conformance or Bronze" caption="Certification judges by per-checkpoint rates and real user tasks; 3.0 conformance judges by all core requirements." >}}
You can't translate "we're certified" into "we conform to 3.0" or "we're around Bronze". Korean certification is a separate scheme based on KWCAG 2.2, and KWCAG 2.2 and WCAG 2.2 share only a version number; they're different standards, so passing certification doesn't mean meeting WCAG 2.2 AA either. WCAG 3 is still a draft and hasn't entered Korean standards. I compared the structural differences between the two standards in [KWCAG 2.2 vs. WCAG 3: What's Actually Different]({{< relref "/posts/kwcag-vs-wcag-3" >}}). That post was written against the March draft and says "WCAG 3 judges by Bronze, Silver, and Gold levels", so please read this post for where the tiers sit.
## What's Still Undecided in WCAG 3 Reporting Tiers, Where Documents Disagree, and Where to Send Comments
First, what hasn't been decided.
- The number of supplemental requirements and assertions to add for Bronze, Silver, and Gold, and whether they're counted separately or together
- The tag for each core requirement and the content or organization tag for each assertion
- How to handle false or outdated assertions, and how to reduce the burden on small organizations (the section above, Two Lingering Worries About Assertions)
- Whether assertions will be required for conformance (editor's note in 4.1.3: "The AGWG has not yet decided whether assertions can be required for conformance.")
- Whether to use the tag-based tiers or the scoring alternative
- The conformance model itself (chapter 3 is still Developing) and the threshold values of some core requirements
If you read the source across several documents, you'll also find places where they describe the same thing a little differently. That's what happens with an exploratory draft, but if you quote it, it's safer to say which section of which document you mean. Unless you're quoting the source directly, feel free to skip this list.
- In 4.1.4 of the September draft, Gold is the tier that adds organization assertions on top of Silver, and the WAI introduction says the same thing if you read it as cumulative (it uses only the tier numbers, not the Bronze, Silver, and Gold names). The scoring alternative in Explainer 7.1.1.3 describes Gold as "many more supplemental requirements and assertions", and doesn't separate out organization assertions.
- Explainer 7 says in one paragraph that tiers above conformance add assertions and recommended practices, and in the next that they're set by supplemental requirements and assertions.
- For what an assertion should include, Explainer 5.3.2 lists six items (what's asserted, date of the statement, date or period the procedure was completed, scope, contact, and the requirements it supports), and the draft's Captions usability testing example lists three items to publish. The example's statement already carries the content, so a straight comparison is hard, but how the two lists fit together hasn't been worked out yet. Explainer 5.3.1 also describes assertions as supporting conformance, which differs slightly from the framing in draft 4.1.3.
The channels for comment are open. The three public comment questions in the WAI introduction ask about exactly this topic: whether a single level plus tag-based reporting tiers is better than previous drafts, what the pros and cons are compared with the scoring alternative, and how you'd use reporting tiers in your own work. You can send comments as [issues on the WCAG 3 GitHub repository](https://github.com/w3c/wcag3/issues), or email public-agwg-comments@w3.org if GitHub is a hurdle. How reporting tiers would work in Korea's certification practice, which runs sample-based review together with user review by reviewers with disabilities, is a perspective English-speaking participants are poorly placed to supply. You can submit as an individual practitioner, too.
## One-Page Summary
- **Conformance** is the verdict on whether you've met every core requirement in your chosen scope. The result is conformant or not conformant, nothing in between. It's normative (MUST) and lives in chapter 3.
- **Bronze, Silver, and Gold** are reporting tiers 4, 5, and 6, which report how much more you did after conforming. They're non-normative, live in chapter 4, and are marked Exploratory.
- **There are four steps to claim one.** Conform (all core requirements), then supplemental requirements, then assertions (do the work, act on findings where the statement requires it, state it), then claim it. Gold also needs organization assertions, and the draft names no body that reviews tiers.
- **Nobody can claim one today.** The counts for each tier are TBD, assertion tags aren't assigned, and with some core threshold values blank, even conformance can't be confirmed yet. Chapter 4 itself is Exploratory and may change.
- **Two open questions remain on assertions.** Evaluation only checks the statement against the documentation requirements, so how to handle false or outdated statements, and how small organizations can bear the burden, are undecided.
- **What to prepare now:** meet WCAG 2.2, run and record the procedures that would become assertions, check the supplemental list against your site, and record severity in two layers in your evaluation reports.
- **Passing Korean quality certification and 3.0 conformance are different verdicts.** Certification needs at least 95% per checkpoint in expert review and a 100% task success rate in user review; 3.0 conformance needs every core requirement in scope.
---
{{< faq >}}
---
## Other Posts in This Series
Apart from the September draft post, the rest were written against the March draft or earlier ones, so they explain Bronze, Silver, and Gold as conformance levels. For where the tiers sit, please read this post and the September draft post.
- [The Dawn of WCAG 3.0: Why We Need New Guidelines]({{< relref "/posts/wcag-3-era-why-new-guidelines" >}})
- [WCAG 3.0 Structure Anatomy: From Success Criteria to Outcomes]({{< relref "/posts/wcag-3-structure-outcomes" >}})
- [WCAG 3.0 Conformance Model: Moving Beyond A/AA/AAA]({{< relref "/posts/wcag-3-scoring-conformance" >}})
- [Atomic Tests vs. Holistic Tests: A New Testing Approach]({{< relref "/posts/wcag-3-atomic-holistic-tests" >}})
- [Assertions: A New Unit for Accessibility Evaluation]({{< relref "/posts/wcag-3-assertions" >}})
- [WCAG 3.0 Expanded Scope: Beyond the Web]({{< relref "/posts/wcag-3-expanded-scope-beyond-web" >}})
- [WCAG 3 Migration: Do You Need It Yet?]({{< relref "/posts/wcag-3-migration-strategy" >}})
- [WCAG 3 Status Check: A Complete Rundown of the March 2026 Draft]({{< relref "/posts/wcag-3-status-2026" >}})
- [KWCAG 2.2 vs. WCAG 3: What's Actually Different]({{< relref "/posts/kwcag-vs-wcag-3" >}})
- [WCAG 3 Accessibility-Support Sets: Korea's Blank]({{< relref "/posts/wcag-3-accessibility-support-sets" >}})
- [What Changed in the WCAG 3 September 2026 Draft]({{< relref "/posts/wcag-3-draft-september-2026" >}})
## References
- [W3C Accessibility Guidelines (WCAG) 3.0 Working Draft (2026-09-10)](https://www.w3.org/TR/2026/WD-wcag-3.0-20260910/): 1.2, 3.1.1, 3.2, 3.2.1.1, 3.2.2, chapter 4, 4.1, 4.1.3, 4.1.4, Captions usability testing, Clear language review, Captions style guide
- [W3C Accessibility Guidelines (WCAG) 3.0 Working Draft (2026-03-03)](https://www.w3.org/TR/2026/WD-wcag-3.0-20260303/): 3.2.1 Conformance level
- [Explainer for W3C Accessibility Guidelines (WCAG) 3.0 (Group Note Draft, 2026-09-10)](https://www.w3.org/TR/2026/DNOTE-wcag-3.0-explainer-20260910/): 5.3.1, 5.3.2, 6.5, 7, 7.1, 7.1.1.3
- [WCAG 3 Introduction (W3C WAI)](https://www.w3.org/WAI/standards-guidelines/wcag/wcag3-intro/)
- [Web Content Accessibility Guidelines (WCAG) 2.2](https://www.w3.org/TR/WCAG22/): 2.3.1, 2.3.2
- w3c/wcag3 issues: [#473 What if it was 10 years ago? - Testing with users (Assertion)](https://github.com/w3c/wcag3/issues/473), [#634 Assertions Feedback and Follow up Q's](https://github.com/w3c/wcag3/issues/634), [#816 AFixt Comments on the WCAG 3 Conformance Model](https://github.com/w3c/wcag3/issues/816) (checked 2026-10-17)
- NIA (National Information Society Agency), Standard Review Guidelines for Web Accessibility Quality Certification Bodies (December 2024, Korean only): chapter 6.2 Technical review (number of sample pages, reviewer types), chapter 7 Review criteria (pass criteria)
- [Korean Web Content Accessibility Guidelines 2.2, KS X OT0003 (National Radio Research Agency)](https://www.rra.go.kr/ko/reference/kcsList_view.do?nb_seq=5247&cpage=1&nb_type=6&kcs_select0=ALL&kcs_select1=ALL&kcs_select2=%C1%A4%BA%B8%B1%E2%BC%FA&searchCon=nb_name&searchTxt=%C1%A2%B1%D9%BC%BA)
- [Talk: What the WCAG 3.0 Draft Asks of Evaluation Methods (Digital Accessibility Standardization Forum, Service Subcommittee, October 16, 2026)]({{< relref "/talks/2026-nia-wcag3" >}})
---
> **Note**: This post is based on the **W3C Working Draft dated September 10, 2026**. WCAG 3 is still under development, and its content may change before the final Recommendation is published. Please refer to the original W3C documents when you need to make an actual decision.