By Dr. Scott Hollier, CEO, Centre for Accessibility Australia.
When I first joined a W3C member organisation in 2009 and became its Advisory Committee representative, we were very excited about the then-new Web Content Accessibility Guidelines (WCAG) 2.0 standard that was released at the end of 2008. Reflecting on the accessibility journey 16 years later, it’s encouraging to see how far we’ve come. While there’s always a need to be vigilant in improvement, it’s great to see the W3C continuing its work in the Web Accessibility Initiative (WAI) to support people with disability in its online content.
In Australia, it’s been particularly important to reflect on this journey, as before April, our national requirement was still locked into WCAG 2.0, but happily, this year marked our move to the current WCAG 2.2 Level AA requirement. It’s been great to see many policy and legislative frameworks around the world adopt the latest version of the standard.
In terms of W3C work itself, there’s been significant progress in the development of the WCAG 3.0 draft, which, while still several years away from being released, is continuing to make progress.
It’s also been great this year to see more guidance on how the WCAG standard can apply to non-ICT requirements, as there are many aspects where the conceptual guidance of the standard, such as good colour contrast, can be applied more broadly to make a variety of scenarios accessible.
The working group I’m involved in, the Research Questions Task Force, which is part of the Accessible Platform Architecture working group, has been continuing the journey in developing international guidance in the area of Machine Learning and Generative AI.
As with much of our world at the moment, generative AI continues to be getting much of the headlines in emerging technology, but it’s important to also consider the accessibility implications: how good is AI-generated alternative text for images? How good is live captioning? Could we reach a point where agentic AI has a mechanism that lives in our web browser and actions our requests, like booking plane tickets, without us needing to deal with an inaccessible website at all? These are all questions that our preliminary AI work, and the work of many others across W3C WAI, are currently addressing.
However, it’s not AI that has been the most controversial issue faced by W3C this year. The organisation as a whole decided to update its logo, which has seen significant internal debate about its design. As with many brand refreshes, some people love the new look, while others feel it’s not a great fit. As I am legally blind, I’ll leave the discussion to people who have better vision than me, so I’ll step aside from the controversy on this one.
Another highlight was the opportunity for my colleague Chris to attend the Technical Plenary and Advisory Committee (TPAC) conference in Japan in November. Here are Chris’ thoughts on the key discussions:
The AG Working Group meetings included considerable process-focused discussions on accessibility standards. To address outstanding issues with WCAG 2.2, a new version within the WCAG 2 series was discussed, although no definitive conclusion was reached.
The WCAG Backlog Task Force continues to work on resolving issues through non-normative changes to WCAG 2.2, which clarify the existing standard but do not alter the normative wording.
Development of the new major version of WCAG, version 3, continued, with a timeline set for its approval as a W3C Recommendation:
– 2029: Normative text finalised
– 2030: Supporting documentation completed
Breakout sessions included an update on the developing WCAG2Mobile group note, which provides guidance on applying WCAG guidelines to mobile apps.
There was considerable focus on the inclusion of Agentic AI in web browsers—“agentic browsing” and the “Agentic Web”—both in formal presentations and informal discussions among delegates. This technology holds great promise for people with disability, as agents may help simplify complex web tasks.
Standards work is needed to support this emerging technology, such as creating API standards for AI models embedded in the browser and defining tags specifically for use by AI agents (or repurposing existing tags such as ARIA).
Care must be taken to ensure that this technology benefits users and does not create new barriers.
Some work at the conference focused on increasing support for augmentative and alternative communication (AAC)—sets of symbols used by non-verbal individuals to communicate. Native support for these symbols would make the web more accessible for these users.
Thanks, Chris, for sharing your insights on TPAC 2025. As I reflect on the year, it’s clear that AI accessibility and the ongoing development of WCAG 3.0 will certainly be in the mix, but as with all things digital access, it’s the unknown developments that will likely shape the year to come.
As the year draws to a close, I’d like to take this opportunity to thank my colleagues that I’ve worked with in the RQTF throughout 2025, and also acknowledge our appreciation as a W3C member organisation to continue to support the development of accessibility web standards and guidance.