Skip to main content

HULAT at Interacción 2026

Foto de Lourdes Moreno, Raquel Hervás, Virginia Francisco

Advances in Human AI project (Towards Accessible and Personalized Language Models) were presented. Human AI is oriented to address challenges  related to people with disabilities when they use Generative AI tools. 

Moreover, Lourdes Moreno introduced the research work  "Oversight Mechanisms for Cognitive Accessibility in LLM-Based Generative Interaction" done by Lourdes Moreno, Paloma Martínez and Blessing Jerry. We propose a mechanism-oriented view that transforms evidence from empirical user studies, standards-based guidance, cognitive accessibility, and accessible content adaptation into concrete oversight mechanisms: signals, decision triggers (rules), KPIs, and forms of human involvement. The approach is illustrated through three proof-of-concept cases: accessible text generation, an LLM-based agentic workflow with stage-wise routing and verification, and UI generation and interaction support using profiles, constraints, and post-render verification.

+ info about Interacción 2026 available here

HULAT at Interacción 2026