REFLECTION-AI: Augmenting Creativity or Compromising Authenticity? Reflections on Using Generative AI in Audio Education

Kim Fox^*

• American University in Cairo, Cairo, Egypt

Introduction This reflection has been conceived based on my experience over the past three semesters utilizing various genAI tools to produce course teaching material. More importantly, while some educators were disallowing the use of genAI, students in some of my courses were encouraged to take advantage of genAI tools to produce audio-related content. The proliferation of genAI in higher education continues to grow and garner discussion about the best way to utilize the various tools and whether the tools should even be allowed by students (Lee & Lui, 2024). However, in media education, many educators are incorporating genAI tools in their courses as a way of job preparedness for students. The sentiment being that the media industry is heavily exploring and utilizing genAI tools to provide news briefs, to help summarize public meetings, to generate story ideas, to cull databases and more. A similar approach is taking place in audio production. National Public Radio's Planet Money demonstrated what's possible when they created a podcast episode using ChatGPT (Malone et al., 2023). Many AI audio tools are now available, like Adobe Podcasts, that have a one-click option to clean up audio (enhance the audio quality, remove background noise/sound, etc). Though diehard audio snobs might balk at some of the genAI options for audio because it dumbs down the intricacies of audio quality. GenAI tools like ChatGPT can be useful to ideate creative projects as well as to generate scripts, particularly when time is of the essence. Time constraints could also be the reason that some educators have resorted to using genAI to assist with course preparation, like generating syllabi, course assessments and rubrics. As we, educators, consider the inclusion of AI in our courses, we also have to grapple with addressing the authenticity, creativity and human agency in learning and in the creative process. As genAI becomes increasingly integrated into educational and creative workflows, it raises pressing questions about the nature of originality, authorship and the human role in meaning-making. In an audio production course, for instance, when a student uses ChatGPT to generate a script or employs a synthetic voice to narrate a story, where do we locate the "authentic" creative voice? Is it in the prompt, the editing, the idea or the final product? These tools can undoubtedly inspire and assist, but do they risk replacing the developmental processes, imagination, struggle and revision, that are central to creative learning? Moreover, ponder this: what happens to students' sense of agency and identity when AI takes on tasks they are still learning to master? If a genAI tool can generate a compelling script in seconds, will students feel encouraged to hone their own storytelling skills, or will they become passive curators of machine output? Similarly, for educators, does the convenience of AI-generated rubrics and syllabi undermine reflective pedagogical design, or can it be a springboard for more dynamic, student-centered learning experiences? In both education and the broader creative industries, these tools challenge us to reconsider the value we place on the process, not just the product. As we integrate AI into classrooms, it becomes essential to create space for dialogue around these issues, so that students don't just use AI, but also question and understand its implications for their craft, their learning and their future roles as media practitioners. I've written about my pedagogical approaches to teaching audio production (Fox & Ebada, 2022). However, the primary goal of this reflection is to explore the opportunities, challenges and ethical implications of genAI in teaching at the collegiate level, specifically in the area of audio production, and to advocate for a balanced approach. Others have written about audio and AI, including in Koh et al.'s paper. They explore the production of an audio fiction podcast created through human-AI collaboration, utilizing tools like ChatGPT for scripting, Prime Voice AI for voice-overs, Evoke Music for soundtracks and Stable Diffusion for visual branding (2024). It proposes a streamlined podcasting workflow and critically examines both the creative possibilities and limitations of AI-assisted storytelling. Scriptwriting with AI as Creative Collaborator Teaching the mechanics of scriptwriting typically takes a great deal of time, with attention given to character and plot development along with other nuances. I've also had my students work on an audio fiction project in my Introduction to Podcasting course. Typically, I culled the internet for public domain scripts, which was time-consuming and didn't always result in the best options that would be of interest to college students. With ChatGPT, we used the tool in real time in class and asked it to generate a 3-4 minute audio drama script based on our specifications, for example, with a similar theme to A Streetcar Named Desire (a favorite from one of my students), but with an Egyptian spin on it. While the output was a bit stereotypical with names like Cleopatra and a requisite scene with the pyramids in it, the students were both impressed with the quickness of the script generation and the formatting and chuckled at the hokeyness of the stereotypes. After generating and confirming the script, students were given the option to use synthetic voices in the production of their audio drama. Many of them used the free version of ElevenLabs. The students expressed frustration at the inability to get the voices to convey the exact emotions that they thought the delivery called for. Whereas, if they were directing voice actors, they could just share verbal directions (prompts) to get their actors to adjust their delivery. We know that text-to-speech synthetic voices are improving; however, in this context, more tweaking (and patience) was needed and the outcome was still not as desired by the students. Although there was a hiccup in this part of the production process, since the final voices were easily identifiable as AI-generated voices, it doesn't mean that I will never allow students to use synthetic voices again. I believe that there is a place for them in the process and I'm glad that students had the practical experience since it impacts their learning of what's possible with genAI tools. However, this tension between efficiency and expressive depth is not unique to the classroom. Journalist, author and podcaster Evan Ratliff's (2024) practice-based experiment Shell Game, a critically acclaimed podcast created using AI-cloned versions of his own voice, tested the medium's ability to convey intimacy and emotion, qualities often cited as hallmarks of audio storytelling. His cloned voices, though technically accurate, repeatedly defaulted to "hackneyed phrases" and "plastic dialogue," with synthetic laughs that lacked spontaneity. As Ratliff observed, these AI hosts "don't sound like they are listening to understand, or to empathize … they are listening in order to prepare a response," ultimately proving incapable of the unpredictability that underpins authentic humor. His findings parallel my students' critiques of synthetic voices in audio dramas, underscoring that while AI can mimic speech, it struggles to replicate the relational dynamics that make audio production a uniquely human art form. AI and the Classroom: Rethinking Teaching Materials and Labor While students' creative engagement with generative AI tools offers exciting possibilities for storytelling and production, these technologies are also transforming the instructional side of education. Beyond the studio, AI is increasingly present in how educators design and deliver course content, prompting a closer look at its role in teaching and assessment. With a skillfully designed prompt, ChatGPT can be very useful in generating course prep materials like syllabi, course outlines and developing new assessments. The output might still need to be tweaked, but generally speaking, it's a time saver. For my course outlines, I provided it with my teaching days and dates along with any days off like holidays. Then I plugged in the assessments and about how many class sessions I wanted to dedicate to each of the assessments, and it generated a fairly usable outline. In one instance, it even provided a suggested list of podcasts for the Listening Reports assessment and I didn't request that information. Eduaide.AI, which has been around since early 2023, has many more pedagogy planning options such as customizing games, quiz questions, group activities and more. Ideally, the use of genAI to assist with course prep and assessments should be considered a timesaver that would allow the professor/educator to, perhaps, commit more time to personalized student interactions. However, despite these efficiencies and innovations, the adoption of AI in both student and instructor workflows brings forth a set of ethical and pedagogical dilemmas (Kostopolus, 2025). As I encourage my colleagues to embrace genAI in educational spaces, it becomes crucial to interrogate how these tools might reshape our understanding of learning, creativity and intellectual ownership. Main Argument - Ethics, Authorship and the Role of the Educator in an AI Era First, let me address concerns about authenticity, creative ownership and reliance on AI tools. In terms of authenticity, when students use genAI to generate scripts, essays or creative content, it becomes difficult to determine what is truly their own work. This raises the question: Is the final product an authentic representation of the student's voice, skill and intent? There's also the question of how to assess the work. Educators may struggle to assess whether learning outcomes, like critical thinking or storytelling ability, are genuinely being demonstrated if genAI is doing much of the creative labor. In the area of creative ownership, students using AI for assignments may not fully understand the boundaries of intellectual property or how to cite their use of generative tools. This creates gray areas around plagiarism and authorship. Lastly, overuse of AI may lead to skill atrophy or a lack of foundational understanding. Students might become overly dependent on AI to brainstorm, write, edit or produce, bypassing key stages of creative development. In navigating the ethical terrain of generative AI in education, The Manifesto for Teaching and Learning in a Time of Generative AI (2024) offers a powerful framework for reframing the educator's role. It urges teachers to cultivate spaces of care, trust and creative risk-taking even amid technological disruption. This resonates strongly with my experience in audio classrooms, where the integration of AI for scriptwriting or synthetic narration risks collapsing nuanced, embodied storytelling into machine-assisted outputs. The manifesto's emphasis on human creativity as a site of resistance encourages educators not to abandon AI, but to use it provocatively, asking students not just what these tools can do, but what they should do. Such a posture enables a shift from efficiency-driven practices toward ethically grounded experimentation, especially in fields like audio production where voice, tone and narrative intimacy are central. Audio educators can leverage AI to revolutionize creative disciplines by using it as a collaborative tool, one that enhances idea generation, streamlines production workflows and expands access to storytelling resources like multilingual voice synthesis or script generation. At the same time, maintaining rigor means embedding reflective practices into the curriculum, where students analyze their creative choices and interrogate the implications of using AI. This dual approach allows for innovation as well as attention to both media and AI literacy (Bali, 2024), preparing students to navigate the evolving audio landscape responsibly. Concerns about the quality and originality of AI-generated outputs are not unfounded. Media strategist Steven Goldstein (2025) cautions against the rise of "AI slop" in creative industries content that appears polished yet is overly predictable, generic and devoid of nuance. In the context of audio production, he identifies a specific risk of "sonic slop," where AI-generated scripts and voices produce technically competent, but artistically flat results (much like what my students experienced), making conversations sound staged rather than organic. Such outcomes pose significant pedagogical challenges: they can normalize mediocrity, obscure the craft of authentic storytelling and discourage students from experimenting beyond AI's defaults. These risks reinforce the importance of cultivating critical engagement and creative agency, ensuring students view AI as a tool for enhancement rather than a replacement for their own imaginative and editorial contributions. Moving forward, consider having students produce a "Hybrid Audio Project," similar to the Koh et al.'s project (2024). Students could create a short audio piece (3-5 minutes) that combines AI-generated content (e.g., scripts, voice synthesis, etc.) with original human input. As part of the assignment, students could submit a brief creative rationale (500-700 words) explaining their choices, what AI generated, what was created or modified by them and why. This allows audio educators to assess both technical fluency with genAI tools and the student's critical engagement and creative authorship, like how to insert authentic emotions to elicit the kind of intimacy that is heralded in audio content. This approach could be useful in fostering a balance of innovation and intentionality. Conclusion As AI becomes increasingly embedded in the creative industries, it is essential that students not only learn how to use these tools, but also critically interrogate their implications. Encouraging this kind of engagement equips students with the ethical awareness, adaptability and creative confidence needed to navigate the AI-driven future that they will soon be a part of. Audio educators are uniquely positioned to lead conversations around the ethical use of AI by modeling responsible use and encouraging experimentation grounded in critical inquiry and artistic integrity. By embracing emerging technologies while prioritizing creativity and authorship along with media and AI literacy, educators can prepare students to be both skilled practitioners and responsible audio storytellers.