TEACHx 2026

As digital accessibility requirements under WCAG 2.1 AA and ADA Title II become increasingly stringent, higher education institutions face significant challenges in providing compliant audio descriptions (AD) for visual-heavy instructional videos. This session presents the results of a pilot project that established a cost-effective, scalable workflow for creating ADs in educational videos in collaboration with human describers/reviewers and Artificial Intelligence (AI).   
Focused on an online Bachelor’s program, the project implemented an iterative process to develop a hybrid workflow: 1) auditing video content by human reviewers based on instructional reliance on graphics, 2) generating baseline descriptions using AI and by human describers, 3) conducting human reviews to ensure pedagogical alignment, and 4) providing feedback on description for AI training and usability experience. We will discuss the challenges and successes of this hybrid workflow. The collaboration highlights refinement areas: humans corrected AI misalignments of on-screen events, adjusted overly detailed descriptions of temporary visual elements, and ensured context was introduced effectively.  
The AI tool will be introduced with AI-created descriptions for sample videos. The presenters will share visual description guidelines for structuring descriptions of technical visuals, such as diagrams, tables, and software for UI interactions. This presentation offers an actionable roadmap for leveraging AI for sustainable and compliant video accessibility.