Abstract
Generative artificial intelligence (AI) use is rising rapidly among graduate students (Yusuf et al., 2024), yet institutional guidance generally remains focused only on plagiarism and academic misconduct (An et al., 2025). Consequently, doctoral students are often left to decide on their own what “ethical” AI use looks like, with little attention to how AI might responsibly support, rather than replace, scholarly work (Brown & Shelton, 2026). This article proposes a shift from compliance-based rules toward an ethics of allocation (i.e., intentionally deciding which tasks to delegate to AI and which to keep for human composition). In this article, we introduce a five-step continuous framework, the DEPTH Cycle (Draft, Evaluate Tools, Prompt, Test Evidence, and Honest Disclosure), and demonstrate its use through a collaborative autoethnography of two doctoral students in Communication Sciences and Disorders at different stages in the program. Although this article focuses on doctoral students, the framework may offer a shared language for all individuals across college and university campuses (e.g., undergraduate students, faculty, and staff). AI is becoming more embedded in everyday academic life; therefore, all members of college and university communities should adopt ethical AI frameworks and guidelines that move beyond prohibition and into AI-supported learning while still protecting critical thinking and authorial voice (Lee et al., 2025).
Recommended Citation
Smith-McMillion, S., Lawrence-Wellman, A. M., Zapata, M., & Harris, O. (2026). The DEPTH Cycle: A Five-Step Framework for Ethical Generative AI Use in Doctoral Education. The Journal of Advancing Education Practice, 7(2). https://openriver.winona.edu/jaep/vol7/iss2/1
