AI in Higher Education: Navigating Responsibly and Equitably

• AI in higher education is an unavoidable reality that requires responsible, equitable, and critical deployment.
• AI tools often exhibit algorithmic bias, such as ignoring unpaid domestic labor when analyzing women's economic contributions, which students must learn to identify.
• Educators are adopting four distinct approaches: traditionalist resistance, pragmatic integration, covert usage, and transparent collaboration.
• AI is framed as an evolution of pedagogical technology, comparable to past shifts like the introduction of statistical software (R, Python, Stata) in econometrics.

The Challenge of AI Literacy

• Generative AI can hallucinate, misattribute sources, and produce flawed explanations with high confidence.
• Students require "AI literacy" to critically evaluate outputs, understand model limitations, and verify information through triangulation.
• Curricula should include algorithmic awareness, prompting students to question data sources, potential exclusions, and the values embedded in human-designed models.

Potential for Inclusion and Accessibility

• Despite risks of exacerbating inequality, AI offers significant benefits for accessibility.
• Specialized tools improve experiences for learners with impairments:
  • Natural language processing for screen readers.
  • Speech-to-text and text-to-speech systems.
  • Real-time captioning and sonification tools.
  • AI-driven sign language recognition.
• Interactive platforms like Mentimeter and Kahoot enhance participation, allowing for anonymous engagement and real-time visualization of learning.