Artificial Intelligence (AI) is often framed as a transformative approach for improving accessibility, with major technology companies investing considerable resources into AI applications targeting disabled users. This investment in AI for accessibility has many benefits but remains relatively unquestioned. Through a critical discourse analysis of 126 public-facing blog posts and news articles by leading U.S.-based AI companies, our analysis reveals the ways in which technology companies render different modes of disabled participation, bestow agency upon AI as a competent and capable actor, reinforce their role in shaping AI futures, and legitimize the development of AI for accessibility. By examining tech companies’ AI visions alongside Critical Disability Studies scholarship, we discuss concerns with framing AI as a means to “solve” disability-related challenges while sidestepping deeper structural questions about equity, agency, and responsibility.