Artificial Intelligence as Assistive Technology in Higher Education: Possibilities of Inclusion and Accessibility for Students With Disabilities at Ufpa

This article analyzes the possibilities of using artificial intelligence (AI) as assistive technology in higher education, focusing on the inclusion and accessibility of students with disabilities at the Federal University of Pará (UFPA). This qualitative and exploratory study is based on bibliographic and documentary research. Brazilian legal frameworks, international guidelines, institutional documents and recent studies on AI, digital accessibility and inclusive education were examined. The findings indicate that speech recognition, text-to-speech, automated image description, captioning, text simplification and content personalization can reduce communication, information and pedagogical barriers. However, algorithmic errors, bias, poor screen-reader compatibility, costs, technological dependence and privacy risks may produce new forms of exclusion. At UFPA, AI has greater potential when integrated with existing accessibility services, subjected to human validation and developed with the active participation of students with disabilities. Institutional adoption should prioritize universal design, data protection, accessibility assessment and continuous training.