AI literacy requires intentional educational opportunities addressing ethical and functional applications. Authors present findings from two undergraduate cohorts (spring 2024 and 2025) who completed a term paper assignment followed by a student discussion thread. Quantitative and qualitative analyses were undertaken revealing limitations that could interfere with student learning and writing based on AI content (5 of 37 papers had errors and 1 was a hallucination) and AI style (AI used the word “crucial” in 28 of 39 responses). Comparison between 2024 and 2025 cohorts showed shifts in students’ perceptions, moving away from inexperience and fear of plagiarism to understanding AI as a multi-purpose tool that cannot replace critical thinking. Four themes revealed were Experience (familiarity, system improvements), Skepticism (trust, fact checking), Drawbacks (promotes laziness) and Tool (brainstorming, editing). Findings led to a 2026 assignment revision designed to support students’ effective AI prompting and detailed disclosure of AI-human interactions.

Presenters

• Brenda Beverly bbeverly@southalabama.edu, University of South Alabama
• Gina Mejia gmm1922@jagmail.southalabama.edu, University of South Alabama