Jemima Abu argues that accessibility should be built into products from the start. She describes a one-off comparison using an accessibility skillAn AI agent skill is a reusable package of instructions, resources, and tool guidance for performing a bounded kind of work., while explicitly cautioning that the experiment does not establish general cost or speed improvementsEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria..
Jemima Abu recommends defining accessibility requirements, auditing existing codeStatic analysis examines code or files for defects and risks without executing the program being analyzed., using reusable components and instructing agents to follow guidelines. She gives semantic HTML and keyboard operation as examples and warns that prompts cannot guarantee a fully accessible result.
In questions, Jemima Abu criticizes both ignoring accessibility and treating an accessibility overlay as a complete fix.
Watch on YouTube
