This paper presents the results of an innovative experiment integrating Generative AI tools into the final drafting process of an analytical paper assignment in an undergraduate Greek and Roman Mythology course. The project required students to apply two interpretive theories to a selected myth, with specific limitations on the use of AI during various stages of the paper's development. Students were tasked with composing an outline and first draft without AI assistance to ensure that all conceptual content was self-generated. For the final draft, AI tools such as Grammarly or ChatGPT were permitted, but exclusively for improving grammar, style, clarity, and the structural organization of the paper—students were prohibited from using AI to develop original content. An appendix was required in which students documented how AI was used and evaluated its impact on their final submission.

The paper discusses the educational benefits and ethical challenges of allowing AI in student work, emphasizing the importance of transparent guidelines for its use and addressing concerns about the boundaries between legitimate AI assistance and academic integrity. Taking into account student feedback, this contribution sets out to offer insights into how AI can be integrated into humanities coursework while preserving the development of core writing and analytical skills.

Keywords: AI, Humanities, Academic Integrity.