NUR 634 Module 8 Policy Paper Example

Neither Ban Nor Free-for-All: A Generative AI Policy for a First Medical-Surgical Nursing Course

[Student Name]

Southern New Hampshire University

NUR 634: Facilitating Learning and Teaching Innovation in Nursing Education

Module Eight Policy Paper

[Instructor Name]

[Date]

The organization, setting and figures below are a composite written as a model document. No real employer, client, colleague or patient is described.

Neither Ban Nor Free-for-All: A Generative AI Policy for a First Medical-Surgical Nursing Course

Within two years of their public release, generative AI tools became part of how many students write, study and search. Nursing faculty have responded in opposite ways: some ban the tools outright, while others say nothing and hope for the best. Neither serves students, who will practice in health systems that increasingly use AI and who must still learn to reason clinically themselves. This paper proposes a generative AI policy for Riverbend Community College's first medical-surgical course. It argues that the policy should protect the skills each assignment is meant to build, require honest disclosure, redesign assessment where AI undermines it and avoid relying on detection tools that are unreliable and unfair.

The Problem in One Course

Last term, faculty in the course noticed several care plans with fluent, generic prose that did not match the patients students had cared for, and references to journal articles that faculty could not find. When asked, two students said they had used a chatbot to "help with wording" and did not realize the references were invented. The course syllabus said only that work must be the student's own. Faculty disagreed about whether using AI to improve grammar was acceptable, and students received inconsistent messages from different clinical instructors.

What the Tools Do Well and Badly

Generative AI can explain concepts in plain language, generate practice questions, help students organize ideas and improve the clarity of writing. It can also produce confident errors. In a study that asked two versions of a popular chatbot to write short literature reviews on 42 topics, 55% of citations from the older version and 18% from the newer version were fabricated, and many real citations contained substantive errors (Walters & Wilder, 2023). For nursing students, who are learning to evaluate evidence, uncritical use of such output is a risk to learning and, eventually, to patients. A reflection on the tools' impact on nursing education argued that faculty should neither ignore nor simply prohibit them, but should teach students to use them critically and transparently while preserving the development of clinical reasoning (Castonguay et al., 2023).

Why Detection Is Not the Answer

Many institutions have turned to AI detection software. The evidence does not support relying on it. A study of several widely used detectors found that they frequently misclassified essays written by non-native English speakers as AI-generated, while rarely misclassifying essays by native speakers, apparently because simpler vocabulary and sentence structure resemble machine text (Liang et al., 2023). In a community college nursing program with many multilingual students, a policy enforced by detection software would risk accusing the students least likely to have used AI and most harmed by a false accusation. Detector scores will therefore not be used as evidence of misconduct in this course.

The Proposed Policy: Three Tiers

Each assignment in the syllabus will be labeled with one of three tiers. Tier 1, AI-free, applies where the assignment's purpose is to build or demonstrate a skill the student must perform independently, such as the in-class unfolding case, clinical judgment reflections written on the clinical day and exams. No AI use is permitted. Tier 2, AI-assisted with disclosure, applies to written assignments such as the teaching plan, where students may use AI to brainstorm or improve clarity but must complete the clinical reasoning themselves and must attach a short statement describing what tool they used, for what purpose and how they checked the output. Every reference must be verified by the student in the library database, and fabricated references are treated as a serious academic integrity issue regardless of their source. Tier 3, AI-integrated, applies to one assignment designed to build AI literacy: students ask a chatbot to write patient teaching for heart failure, then critique its accuracy, reading level and cultural fit against evidence-based sources.

Redesigning Assessment

Policy alone cannot protect learning if assessments can be completed by AI without thought. The course will shift weight toward work that requires the student's own clinical experience and reasoning: care plans will be replaced by structured reflections tied to specific moments with the student's own patient, and one written assignment will include a short oral discussion in which the student explains his or her reasoning to the instructor. These changes make undisclosed AI use both less useful and easier to notice through conversation rather than software.

Faculty need preparation as well as a policy. A one-hour workshop before the term will let instructors try the tools themselves, see how they handle a nursing prompt and where they fail, and practice the dialogue-first conversation described below. Faculty will also agree on how to respond to common questions, such as whether a grammar checker counts as AI, so that students hear consistent answers. The policy will be reviewed at the end of each term, because the tools change quickly, and student representatives will be invited to comment on whether the tiers are clear and fair. Treating the policy as a living document signals to students that the aim is learning, not policing.

A Fair Process for Concerns

When an instructor suspects undisclosed AI use in a Tier 1 or Tier 2 assignment, the first step will be a conversation with the student about the work, not an accusation. The instructor will ask the student to explain the reasoning and sources. If concerns remain, the case will follow the college's existing academic integrity process, with evidence drawn from the student's explanation, the assignment itself and verification of references, not from detector scores. All clinical instructors will receive the policy and a brief orientation so that students hear one consistent message.

Conclusion

Generative AI is neither a threat to be banned nor a convenience to be ignored in nursing education. A three-tier policy that protects the skills each assignment builds, requires disclosure and reference verification, redesigns vulnerable assessments and avoids unreliable detectors gives Riverbend's students clear expectations and prepares them to use these tools critically in practice.

References

Castonguay, A., Farthing, P., Davies, S., Vogelsang, L., Kleib, M., Risling, T., & Green, N. (2023). Revolutionizing nursing education through AI integration: A reflection on the disruptive impact of ChatGPT. Nurse Education Today, 129, Article 105916. https://doi.org/10.1016/j.nedt.2023.105916

Liang, W., Yuksekgonul, M., Mao, Y., Wu, E., & Zou, J. (2023). GPT detectors are biased against non-native English writers. Patterns, 4(7), Article 100779. https://doi.org/10.1016/j.patter.2023.100779

Walters, W. H., & Wilder, E. I. (2023). Fabrication and errors in the bibliographic citations generated by ChatGPT. Scientific Reports, 13, Article 14045. https://doi.org/10.1038/s41598-023-41032-5