Accessibility is not a polish pass — what this month’s standards news should change
15 October 2026 / Always 49
In early September 2026, a new version of Europe’s core ICT accessibility standard, EN 301 549 v4.1.1, was published. The headline for anyone who builds websites and software is straightforward. The technical benchmark is moving from WCAG 2.1 thinking toward WCAG 2.2. Focus appearance, target size, dragging movements, accessible authentication, and a stronger respect for user preferences are no longer “nice extras for later”. They are the direction of travel, including for organisations that sell into Europe or simply do not want to build a product that excludes people.
We are a Bristol studio, not an EU compliance house, and we will not pretend otherwise. We still think this news should change how UK teams plan work. Accessibility has a habit of being treated as a national paperwork problem. Users do not experience paperwork. They experience whether they can see the focus, whether they can tap the button, whether they can log in without a puzzle that only works for people with perfect memory and perfect vision. Those people are in your user base whether your lawyer has mentioned the EAA or not.
At Always 49 our opinion is blunt. If accessibility is a phase after the “real” design, you do not have a UX-first process. You have a visual process with a guilt extra. Accessible authentication, for example, is not a sticker you put on a login screen. It is a product decision about passwords, magic links, copy-paste, password managers, and whether you have forced people to remember a code they cannot see. Get it wrong and you have locked someone out of a service you claimed was for everyone. That is not a polish defect. That is a broken promise.
The same is true of target size and focus. We still see interfaces that look refined in a screenshot and collapse under a keyboard or a thumb. Refinement that only works with a mouse is not refinement. It is a narrow bet on one kind of body and one kind of device. Design for extremes again: the person using voice control, the person on a bus, the person with a temporary injury, the person who never noticed they could tab through a form. When those paths work, the software feels calmer for everyone.
There is a project-management lesson hiding in the standard update, and it is the one we care about as much as the criteria themselves. Teams that bolt accessibility on at the end pay twice. They pay to build it wrong, then they pay to unpick components, design tokens, and authentication flows that were never set up to flex. We would rather put the constraints in Discover and Design, then develop against them. Constraints are not the enemy of creativity. They are how you stop inventing decorative interactions that nobody can operate.
AI makes the timing of this news more pointed, not less. Generated interfaces can look WCAG-ish while failing the things automated checkers are weak at: focus order, name, role, state, and whether a drag-only action has a keyboard equivalent. If your team is using assistants to draft UI, you need a human accessibility pass that is allowed to fail the sprint. Otherwise you are scaling the inaccessible average of the web at new speed. That is the opposite of user-first software.
We also think “conformance” can become a trap if it replaces empathy. A checklist can get you a long way, and you should use one. A checklist cannot tell you whether the help text is kind, whether the error is usable, whether a person in a difficult moment can recover. Our process still wants a real person in the loop, including people who use assistive technology, not only a dashboard of automated scores. Scores are a floor. Experience is the point.
If you are planning a product this autumn, do not wait for every legal citation to catch up before you take WCAG 2.2 seriously. The users who needed larger targets and visible focus last year still need them this year. Always 49 will keep treating accessibility as part of designing for simplicity and effortless control, not as a certificate to chase after the screens are frozen. Standards news is useful. It is even more useful if it changes the order of the work.