Artificial intelligence (AI) holds enormous potential to improve the accessibility of digital systems. We hypothesize that within the digital accessibility ecosystem, students in the field of interaction design (as prospective practitioners) will play a key role in adopting AI-powered solutions. Understanding their current opinions and perspectives on these technologies is crucial to anticipating the future adoption of AI to create more accessible systems. Therefore, we conducted an exploratory survey to investigate students’ perspectives on the use of AI in digital accessibility. We also examined their beliefs on the related risks and ethical considerations. Study results showed that most participants identified AI’s potential to enhance existing assistive technology and improve the effectiveness of current tools. A primary issue that most participants raised revolved around concerns about accuracy and reliability. The risk of overreliance, the lack of empathy, and human oversight in AI integration were also prominent. Participants generally recommended the most obvious practical solutions for using AI for accessibility, leading us to discuss the importance of awareness, AI literacy, and education, as well as the practical application of knowledge to promote it. So that future developers can leverage the full spectrum of accessibility improvements that AI promises. This study contributes insights into those perspectives and analyzes how those would affect the adoption of AI.