JHDD UI UX Report — 2026.09.01

Google’s A2UI project suggests a future where software agents compose interfaces from trusted component catalogs.

This development, alongside the burgeoning interest in AI for UX and the re-evaluation of design careers, points to a fundamental shift in design’s core concerns. The emerging pattern is an increasing abstraction of interaction itself, moving beyond fixed interfaces to dynamic, interpreted user goals. Design is less about crafting pixel-perfect layouts and more about defining the parameters, logic, and ethical boundaries for intelligent systems that generate their own interactions and responses. This challenges long-held principles of direct manipulation and traditional craft, demanding a different kind of expertise from professionals.

The design industry often views Jakob Nielsen’s sixth usability heuristic, “recognition rather than recall,” as immutable guidance for crafting visible, persistent elements. This perspective, however, overlooks the deeper implications of agentic interfaces, such as Google Search generating interactive graphs or mini apps in response to queries. Conventional wisdom holds that a designer’s primary task is to minimize memory load by meticulously arranging elements for direct user recognition, ensuring “nobody wants a blank page.” A contrasting view posits that in an era of adaptive systems, the system will increasingly handle the recognition and recall, interpreting user intent to dynamically surface the most relevant interaction or information. This fundamentally redefines “visibility” and the problem of the blank page. The challenge is not to avoid blank pages by pre-designing every element, but to empower systems to generate the appropriate interaction on demand, often without requiring the user to navigate a predefined structure. While courses like Designlab’s AI for UX Design teach integrating AI into existing workflows, the deeper shift implies that the ‘workflow’ itself must adapt to this agentic reality.

This shift also affects how design systems are conceived and managed. Traditionally, design systems provide rigid, reusable components primarily for human designers. With Google’s A2UI project, agents are now consuming these catalogs to compose interfaces. Gartner already predicts task-specific agents appearing in 40% of enterprise applications by the end of 2026. The implication is profound: design systems become less about visual consistency enforced by human designers, and more about functional consistency and interoperability for intelligent agents. The mainstream focus on design systems as a static library of UI elements for human consumption will prove insufficient. Instead, design systems will evolve to include semantic metadata, behavioral parameters, and trust-level classifications that agents can interpret. Within two years, organizations will begin investing heavily in “agent-ready” design systems, prioritizing machine-readable attributes over purely visual specifications, making them a crucial layer in AI governance and interaction.

The primary resistance to this evolution comes from established corporate structures and ingrained design education models. The “endless staircase for endless metrics” described in career discussions, with its emphasis on linear progression and quantifiable output, creates an organizational inertia that often struggles to value abstract, system-level design contributions. These environments prefer metrics tied to traditional interface delivery, rather than rewarding expertise in defining agent behavior, crafting prompt taxonomies, or auditing AI outputs for usability and bias. The mental model of design as purely visual craft, as reflected in many traditional “September wallpapers” and the focus on aesthetics, still holds significant sway. This creates a gap between emerging practices and institutional recognition, making it harder for designers who embrace neuroqueering principles—challenging dominant narratives and seeking guidance beyond craft expertise—to thrive in conventional settings.

A working UI UX professional should this week begin to explore and experiment with defining functional parameters and semantic tags for existing design system components, rather than solely focusing on their visual attributes. Practical engagement includes experimenting with prompt engineering techniques to generate interfaces or interactions using AI tools, then critically evaluating the output against established usability and accessibility heuristics. This means learning to “design for the agent” as much as “design for the human,” understanding how AI interprets and constructs interactions, and treating AI-generated elements as inputs to be refined and governed rather than finished products.

---

Curated References