Kate Agena argues that AI experiences cut across the organizational boundaries that traditionally separate product design, content, knowledge management, engineering, and governance. She is right that these systems require specialized expertise. But AI may also change who can apply that expertise.

A well-designed AI framework can encode the methods, standards, and questions that specialists use. This does not transform a novice into an expert. It gives more people structured access to expert practice.

I am already seeing this in design work.

An AI-assisted workflow can use an established design system to check whether interface decisions remain consistent. Specialized roles can examine responsive behavior, color contrast, accessibility requirements, interaction patterns, and content clarity. Each role applies a different framework, while sharing the same product context.

The model is not one apprentice learning from a master. It is closer to many apprentices learning the master’s methods and then taking on specialized responsibilities. One examines accessibility. Another checks responsive behavior. Another looks for design-system inconsistencies. Another evaluates the experience from the perspective of a defined persona.

Together, they can make specialized review more consistent and available earlier in the design process.

But the framework is only as good as its source material and its human oversight.

An AI-generated accessibility audit is not a completed accessibility assessment. Automated contrast checks cannot determine whether an interaction is understandable, whether assistive-technology users can complete a task, or whether the experience works under real conditions.

The same limitation applies to simulated personas. Asking AI to perform a task from the perspective of a persona can expose assumptions and generate useful questions. It is not evidence that real people will behave that way. Synthetic users should help teams prepare for research, not replace contact with actual users.

This is why every AI-assisted workflow needs a human gatekeeper. The human reviews the findings, challenges unsupported conclusions, refines the work, and decides which assumptions require validation with real users.

Consistency and context are the real opportunity. AI can help teams apply specialized frameworks repeatedly without forgetting the standards, dependencies, or user conditions that matter. Human expertise still determines whether those frameworks are appropriate and whether the resulting recommendations deserve to be trusted.

AI does not eliminate specialization. It can distribute the practices behind specialization, giving more people the ability to produce informed work while keeping accountability with the humans qualified to judge it.

Source

Kate Agena, PhD, “The AI experience doesn’t care about your org chart”, UX Collective, October 15, 2026.