Accessibility Challenges in an AI-Driven World

Описание: In this episode, Rashmika speaks with Oriol Gómez, Accessibility Lead at Dow Jones, Barcelona, about building accessible products in an AI-driven world and how AI is reshaping workflows for developers, designers, and people with disabilities.

This episode is perfect for:
Designers, developers, accessibility specialists, AI builders, and anyone curious about how AI intersects with inclusive design.

Host:   / rashmikaprasad  
Guest:   / ogomez92  

https://github.com/ogomez92/descam_web
https://github.com/ogomez92/eleven_la...
https://github.com/ogomez92/arm


Summary
In this insightful conversation, Oriol shares his journey from translator to programmer to accessibility leader, highlighting how AI has empowered his own workflows as someone with a visual disability. He explains where AI supports accessibility, where it still fails, and why empathy, foundational accessibility knowledge, and intentional prompting matter more than ever.

The episode covers the future of accessibility standards, designing AI-powered products, keyboard-friendly interaction patterns, the challenges of coding with AI tools, and the importance of human-centric decision-making.

Frameworks Discussed
• Accessibility-First Development Mindset
– Build accessibility at the foundational stage
– Never rely solely on user feedback to fix accessibility debt
– Use AI to support coding and documentation, not to replace expertise
• Prompt Engineering for Accessibility
– Always specify: accessibility requirements, semantic HTML, ARIA needs
– Guide AI with step-by-step instructions for complex interactions
– Review all AI-generated code with an accessibility lens
• Keyboard Interaction Model
– Provide alternatives for drag-and-drop
– Ensure tooltips, lists, and flows work fully via keyboard
– Designers must communicate interaction requirements early to developers

AI Tools Mentioned
• ChatGPT – general workflows, image descriptions
• Gemini – image descriptions and accessible Figma component descriptions
• Claude – accessibility challenges due to unlabeled UI elements
• Suno – AI music generation used for fun and creative exploration
• AI Coding Tools (GPT-5, Sonnet, Codex-style models) – used for “vibe coding,” with limitations
• Figma’s upcoming AI features (mentioned in context of design systems)

Important Highlights
• How AI helps accessibility professionals become more self-reliant
• Why prompt engineering is essential for generating accessible code
• The gap between accessibility testing tools and real-world needs
• The risk of “vibe coding” and why understanding code is still crucial
• Why accessibility must be built from the beginning, not added later
• How big SaaS tools (Slack, Salesforce, etc.) evolve accessibility over time
• What designers must understand about drag-and-drop, keyboard flows, and alternative interactions
• The “human superpowers” we must sharpen to stay relevant in the AI era
• The danger of AI mixing legacy and modern code when not guided well
• Why empathy and user understanding will always outperform automation

If you enjoyed this episode, like, comment, and subscribe for more conversations on AI, design thinking, and the future of user experience.

Editing Credits: Param ([email protected])

#ai #accessibilitystandards #inclusivedesign #a11y #uxdesign #productdesign #aiindesign #WebAccessibilityAI #accessibletechnology #assistivetechnology #designsystems #promptengineering #aiworkflows #techforgood #digitalaccessibility #AccessibleCoding #EmpathyDrivenDesign #saasproducts #AIProductDesign #humancentereddesign #FigmaAI #aidevelopment #UXPodcast #designpodcast #aiinterviews #dowjones #AccessibilityLead #sunoai #gpt4turbo #geminiai #claudeai

Reach out to the host!
LinkedIn:   / rashmikaprasad  
Medium:   / rashmikaprasad26