LLMs Generate Accessibility-Aligned Multimodal User Interfaces

Researchers led by Lourdes Moreno on Jan 10, 2026 present a model-driven architecture that uses LLMs to generate personalized, multimodal, accessibility-aligned user interfaces. It combines structured user profiles, declarative adaptation rules, and validated prompt templates to produce Plain-Language text, pictograms, and high-contrast layouts compliant with WCAG 2.2, EN 301 549, and ISO 24495-1. A healthcare use case demonstrates explainable SysML v2 traceability.