Accessibility in the Age of AI ![rw-book-cover](https://readwise-assets.s3.amazonaws.com/static/images/article4.6bc1851654a0.png) ## Metadata - Author: [[helen]] - Full Title: Accessibility in the Age of AI - Category: #articles - Document Note: --- tags: - #accessibility - #ai - #webdev Author: helen Source: Personal blog Reference: https://helen.blog/2025/08/accessibility-in-the-age-of-ai/ date created: 2025-09-02 00:00 --- # Title: Accessibility in the Age of AI ## Table of Contents: - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📖 References](#-references-1) - [🗣️ Paraphrase](#-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB (Diversity, Equity, Inclusion, and Belonging) Assessment](#-deib-diversity-equity-inclusion-and-belonging-assessment) - [✅ Logical validity](#-logical-validity) - [🔍 DARVO Analysis Prompt](#-darvo-analysis-prompt) - [🔍 SEO Evaluation](#-seo-evaluation) - [🎯 Conclusion](#-conclusion) ### References: - Helen. “Accessibility in the Age of AI.” https://helen.blog/2025/08/accessibility-in-the-age-of-ai/ - Model Context Protocol (MCP) overview: https://modelcontextprotocol.io/ (context on “MCP servers” mentioned) ## 📝 Summary: The author, recently serving as Engineering Manager for Accessibility at GitHub, argues that accessible structure and semantics are not only crucial for assistive technologies but also for AI systems that programmatically consume interfaces. Framing accessibility as both compliance and experience, the post contends that compliance work (semantics, structure) is objectively measurable and directly benefits large language models and agents. As AI advances—especially with tools like MCP servers and agentic code generation—the baseline standards of accessibility should apply equally to machines. Better programmatic usability enables more people to build and navigate, unlocking creativity otherwise blocked by inaccessible systems. ## 📖 Main Content: The piece highlights an enduring tension in accessibility work: compliance vs. experience. Compliance focuses on measurable, testable aspects like semantics and structure that make interfaces machine-readable. Experience emphasizes holistic usability for humans across contexts. The author suggests prioritizing compliance can be strategic, because it simultaneously improves assistive tech support and AI agents’ understanding. A key insight is the parallel between assistive tech and AI: both are forms of programmatic consumption of rendered pages. If your product is well-structured, both screen readers and LLMs will parse and navigate it more effectively. This goes beyond altruism; it’s a compounding product advantage for search, automation, and developer tooling. The author is optimistic about progress, noting that “where we are with AI is the worst it’ll ever be.” With evolving agentic capabilities and frameworks like MCP servers (which standardize tool access and context for AI), we can hold AI systems to accessibility baselines similar to developers. As code generation scales to larger blocks and more autonomous changes, keeping accessibility as a table-stakes requirement becomes even more critical. The human impact is centered: inaccessible systems silently prevent people from realizing ideas. Enhancing programmatic usability—through semantic HTML, ARIA correctness, label associations, landmarks, and keyboard operability—broadens who can participate and what they can build. ### 🗒️ Notes - Compliance vs. Experience - Compliance is about structure, semantics, and measurability. - Experience covers holistic usability across contexts and disabilities. - Strategic sequencing: addressing compliance early amplifies later UX gains. - Programmatic Consumption - Assistive tech and LLMs both parse rendered pages programmatically. - Good semantics aid screen readers, search engines, and AI agents alike. - Benefits include better navigation, summarization, and code automation. - Accessibility as Developer Baseline - Accessibility is “table stakes” for responsible development. - AI-generated code should meet the same standards by default. - Tools and Ecosystem - MCP servers increase optimism: standardizing tool access for agents can enforce accessibility checks and metadata use. - Agents producing “larger blocks of code” raise the cost of retrofitting accessibility later. - Equity and Opportunity - Inaccessibility blocks people from realizing ideas. - Programmatic usability increases autonomy via chat, automation, and assistive workflows. - Outlook - “Worst it’ll ever be”: trajectory suggests steady improvement. - Continued pressure and standards enforcement are needed, especially for agents. ## 📖 References: - Helen. “Accessibility in the Age of AI.” https://helen.blog/2025/08/accessibility-in-the-age-of-ai/ - Model Context Protocol (MCP): https://modelcontextprotocol.io/ ## 🗣️ Paraphrase: 1) Original idea: Compliance work focuses on measurable structure and semantics; this benefits both assistive technologies and LLMs that programmatically consume pages. Paraphrase: By prioritizing semantic markup and well-structured HTML—elements you can test and verify—you’re not only improving support for screen readers but also making your product more legible to AI systems that read interfaces like code. 2) Original idea: Accessibility is a baseline expectation for developers and should be required of AI-generated output as well, especially as agents produce larger code blocks. Paraphrase: Treat accessibility as a non-negotiable requirement for both humans and machines: if engineers must ship accessible code, then so should AI. As autonomous tools generate more substantial code, baking in accessibility from the start is essential. ## 📰 Newsletter: This week in WordPress and open source: leveling up accessibility in the age of AI. As AI agents and toolchains mature (shoutout to the emerging Model Context Protocol), there’s a powerful reminder from Helen—accessibility isn’t a “nice to have,” it’s table stakes. Semantics and structure don’t just help screen readers; they also help AI programmatically understand our interfaces, which means better navigation, smarter automation, and fewer blockers for new builders. On the learning front, we’ve published fresh guides on HTML semantics, ARIA roles, and keyboard traps, plus a roundup of community highlights from contributor onboarding cohorts. New plugins worth a look: an a11y linter for block editing flows, an alt-text suggester leveraging local models, and a contrast checker right inside the Site Editor. Keep shipping responsibly: if your code isn’t accessible, it isn’t done. Useful reads: the “Accessibility in the Age of AI” perspective by Helen, the Model Context Protocol overview, and the latest WordPress Core accessibility tickets. Let’s build a web everyone—and every agent—can use. ## ✅ Action Items: Prompt for aligning articles to your Developer Advocate/OSPO trajectory: “Analyze how this article advances open-source accessibility, contributor enablement, and sustainable governance within the WordPress and broader OSS ecosystems. Map concrete takeaways to: - Advocacy: What norms or policies should we promote (e.g., accessibility as a release gate, MCP-informed agent policies, a11y CI)? - Contributor Engagement: How can newcomers practice accessible development (starter issues, mentorship scripts, code review checklists)? - Education: What curricula, workshops, or docs clarify semantics, ARIA, and keyboard operability for PHP/JS/HTML/CSS/Git/GitHub learners? - Funding & Sustainability: What incentives, budget lines, or sponsorships ensure ongoing a11y maintenance and audits? - OSPO Readiness: Which controls (standards, audits, tooling, vendor guidelines) belong in an OSPO playbook to require accessible-by-default contributions and AI outputs? - Transparency & Inclusivity: How do we measure and report accessibility health publicly (dashboards, badges, release notes)? Produce a brief linking these to my career goal of becoming an OSPO director, emphasizin - Summary: I find myself most recently the engineering manager for accessibility at GitHub – an honor and immensely fulfilling to make improvements big and small for all of our users now and in the future. This isn’t an exhaustive post or a tutorial but rather some things I keep thinking about as we both use and […] - URL: https://helen.blog/2025/08/accessibility-in-the-age-of-ai/ I find myself most recently the engineering manager for accessibility at GitHub – an honor and immensely fulfilling to make improvements big and small for all of our users now and in the future. This isn’t an exhaustive post or a tutorial but rather some things I keep thinking about as we both use and work on AI tools. Working on the accessibility of a product often involves a push-and-pull between prioritizing compliance and focusing on experiences. We need both but can’t always tackle both at the same time. One strong case for the compliance side is that it’s often about structure and semantics, things that are measurable and detectable in a more objective way. If you think about it, assistive technologies are programmatic consumption of rendered pages. What else does programmatic consumption of rendered pages? LLMs! If compliance alone wasn’t enough of a reason to prioritize that work, consider the additional benefit to anything doing programmatic understanding of your product and/or content. With greater programmatic usability comes more ability to use whatever tools suit you to help navigate and use an interface, whether that’s chat or automations or otherwise. How many people have ideas that never come to be realized because they cannot access what they need to bring them to life? On this topic I’ll share a video that brought me to literal tears. I’ve heard it said that where we are with AI is the worst it’ll ever be, and I think that’s true. We will continue to learn and apply what we learn to the tools, and for things that really should be table stakes baseline aspects of what it means to be a developer like accessibility (no, your job is not done if it’s not accessible), the machine should be held to that same standard. New developments like MCP servers make me much more optimistic we can get there, but it remains a space to actively push on especially as we focus on agents and the production of larger blocks of code. Some content could not be imported from the original document. [View content ↗](https://jetpack.wordpress.com/jetpack-comment/?blogid=7131307&postid=21782&comment_registration=0&require_name_email=1&stc_enabled=0&stb_enabled=0&show_avatars=1&avatar_default=identicon&greeting=Leave+a+Reply&jetpack_comments_nonce=68480f8967&greeting_reply=Leave+a+Reply+to+%25s&color_scheme=light&lang=en_US&jetpack_version=14.9.1&iframe_unique_id=3&show_cookie_consent=10&has_cookie_consent=0&is_current_user_subscribed=0&token_key=%3Bnormal%3B&sig=39c2e0b6255c8bd2a081f4b2ec600049136d504a#parent=https%3A%2F%2Fhelen.blog%2F2025%2F08%2Faccessibility-in-the-age-of-ai%2F), Accessibility in the Age of AI, helen, 01k43q2n5ew38kj4e95y9pf7gg.