Statement on Artificial Intelligence

Statement on Artificial Intelligence

First-Year Writing Program

Emory Writing Program

Spring 2026

In the First-Year Writing Program at Emory University, students (1) develop writing processes that include phases of brainstorming, outlining, revising, and editing; (2) target their compositions to specific audiences and analyze the affordances of specific modes of communication (e.g., oral presentations, podcasts, and research posters); and (3) use critical reading and thinking as a means to develop writing. 

Each of these outcomes links the growth of a student’s compositional abilities to the growth of their intellectual abilities. To determine a writing process, analyze an audience’s interests, assess a mode of communication, or contribute meaningfully a complex discourse, students must reflect on their own experiences with writing, in writing, to consider the nuances of the texts they encounter, and to develop original provocations that push existing discourses into new territory. As students learn to do so in First-Year Writing, they build a foundation for success in future discipline-specific courses. Equally important, they discover and sharpen the tools needed to live considered lives, thus realizing Emory’s university mission.ⁱ

Artificial Intelligence software threatens students’ opportunities to develop this essential foundation. To illustrate, a study hosted on Microsoft.com found that “higher confidence in GenAI is associated with less critical thinking.”ⁱⁱ. Relatedly, a study led by a researcher at the Samuel DuBois Cook Center on Social Equity at Duke University observed, in the context of reading and writing based tasks, “a decrease in learning effectiveness in both AI-based and AI-assisted scenarios, with declines of 25.1% [AI-based] and 12% [AI-assisted] respectively.”ⁱⁱⁱ By AI software, we mean generative, prompt-based text or image platforms that use large-language-model prediction software, trained on millions of people's stolen intellectual property, to predict sequences of text or provide images in response to a human generated prompt. Furthermore, this statement refers to large-scale commercialized models of this software such as ChatGPT, Claude, Co-Pilot, Grammerly, and others. 

Uses of GenAI platforms as a “reading resource” or “thinking partner” further contradict the deployments of attention and sustained process-based writing on which academic thinking depends. For example, when scholars read primary and secondary sources, they rarely read only for “main points”; they instead respond attentively to both the large arguments and small details. An AI summary of a text is likely to deny students the chance to identify and engage with those small details and see how seemingly minor points and textual details raise necessary questions or provoke sophisticated responses. Similarly, a predictive-text model trained on preexisting instances of writing cannot direct students to new intellectual ground; it can only retread ground that has already been explored. Generative AI platforms cannot, by design, have new thoughts. As such, they cannot invite students to develop their own new knowledge. 

While AI is certainly enmeshed in many facets of professional and even scholarly life, the studies listed here, and many more, suggest that the most effective users of a range of AI platforms will not be those who acclimate to the software the earliest. Instead, and as the Duke University study corroborates, the most effective users will be those whose use follows from a rigorously developed practice in independent thinking, reading, and expression.^iv 

For this reason, the First-Year Writing Program endorses and follows Kathyrn Conrad’s directive in “A Blueprint for an AI Bill of Rights for Education” to communicate expectations around AI use or prohibition to students clearly (and prior to the assignment of any assessed work).^v 

Students and colleagues can view the Emory Writing Program's GenAI policy for classroom and instructor expectations as a separate document.

It is all but impossible for university students to avoid genAI platforms during their studies. The First-Year Writing Program takes this fact as a reason to empower instructors to create learning environments in which students can read, think, and collaborate in a manner befitting the development of their most essential capacities.

i “Mission and Principles.” Emory University. Accessed Feb. 25, 2026. https://www.emory.edu/home/explore/emory-profile/our-mission.html 

ii Hao-Ping (Hank) Lee, Advait Sarkar, Lev Tankelevitch, Ian Drosos, Sean Rintel, Richard Banks, and Nicholas Wilson.” The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects from a Survey of Knowledge Workers.” Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI ‘25), Association for Computing Machinery (2025): 1-22. https://doi.org/10.1145/3706598.3713778

iii Ju, Qirui. “Experimental Evidence on Negative Impact of Generative AI on Scientific Learning Outcomes.” arXiv preprint(2023). http://dx.doi.org/10.2139/ssrn.4567696

iv Ju also finds that “without mastering the knowledge through rigorous study and engagement, students become more susceptible to accepting and propagating false information.”

v Conrad, Kathryn. “A Blueprint for an AI Bill of Rights for Education.” Critical AI vol. 2, no. 1 (2024). https://doi.org/10.1215/2834703X-11205245