ai-in-nursing-education-critical-thinking

AI in Nursing Education: How Faculty Can Embrace Technology Without Sacrificing Critical Thinking

Share

Artificial intelligence can generate patient scenarios, develop practice questions, and help nursing students review complex clinical concepts in seconds. But as AI becomes more common in nursing education, faculty face a critical challenge: How can they use the technology without letting it do the thinking for their students?

The stakes extend beyond the classroom. Nurses must recognize subtle changes in patient conditions, prioritize interventions, and make clinical decisions based on evidence and individual patient needs. An AI-generated answer may sound convincing, but it may not be clinically appropriate or safe.

As National Nurse Practitioner Week, Nov. 8–14, recognizes the contributions of advanced practice nurses, nursing educators are exploring how AI-supported learning can strengthen the clinical judgment students will need throughout their careers.

Using AI to Create More Realistic Clinical Learning Experiences

At Purdue Global School of Nursing, Abbey Elliott, associate dean of innovation and wellness and coordinator of immersive learning, describes an approach that positions AI as a faculty-guided practice partner rather than a shortcut.

Elliott describes an AI-Assisted Patient Analysis & Critique activity in which students evaluate an evolving patient scenario involving fluid overload or respiratory distress.

They enter baseline assessment information into an AI tool and ask it to identify three nursing priorities and develop an intervention plan.

The assignment doesn’t end when AI generates its recommendations.

Students must evaluate the proposed plan against evidence-based references, clinical assessment criteria, and electronic health record learning tools, such as Lippincott DocuCare.

Their written analysis must identify at least one subtle clinical cue the AI overlooked, verify medication and dosage appropriateness, and explain why their revised care plan is safer for the patient.

The exercise turns AI-generated information into material students must question rather than accept.

Elliott also describes using AI for personalized concept review, simulation preparation, clinical scenario development, and guided reflection.

These activities give students opportunities to explore clinical decisions before entering simulation laboratories or patient care environments.

Helping Students Move From Memorization to Clinical Judgment

Lisa M. Murphy, MSN, RN, a doctoral candidate and nursing faculty member at Maria College, uses AI to develop learning activities that challenge students to apply clinical concepts rather than recall information.

Murphy teaches through a concept-based, flipped-classroom approach and uses ChatGPT and Google Gemini to help create unfolding patient scenarios, Next Generation NCLEX-style questions, concept maps, pre-class worksheets, and interactive review activities.

She often uploads course presentations and learning objectives into Gemini, instructing the tool to generate supplemental materials using the information she provides.

The approach helps align activities with course content, although she emphasizes that restricting AI to supplied materials does not eliminate errors.

Murphy reviews and revises generated content for clinical accuracy, appropriate complexity, and alignment with learning objectives.

Her scenarios require students to recognize important assessment findings, prioritize interventions, anticipate provider orders, and determine appropriate next steps.

She also uses the National Council of State Boards of Nursing’s Clinical Judgment Measurement Model to structure activities around clinical reasoning.

By generating different versions of patient scenarios, Murphy challenges students to apply their knowledge to changing circumstances rather than memorize familiar answers.

Teaching Students to Think Before Consulting AI

One concern for nursing educators is that students may become accustomed to immediate answers without developing the reasoning needed to reach them independently.

Elliott addresses this through what she calls the Attempt Without Rescue principle.

Before consulting AI during clinical simulations or case analyses, students must document their initial concern, identify important assessment cues, and commit to a clinical action.

See also  NP Week 2026: Free On-Demand Webinars for Nurse Practitioners

Only then can they use AI to evaluate their thinking.

Assignments also require students to submit their original prompts, AI-generated responses, and annotations identifying weaknesses or necessary corrections.

Juliet Kolde, PhD, RN, CNE, president of Nightingale Innovations at Nightingale Education Group, emphasizes teaching students to recognize when AI is helpful and when they should rely on their own judgment.

At Nightingale College, educators are working to integrate AI literacy throughout the curriculum rather than treating responsible technology use as a one-time lesson.

Kolde recommends a comparison-based assignment in which students complete a task independently, repeat it with AI assistance, and analyze the differences.

Students must explain which approach is stronger and why.

As students advance, Kolde recommends assignments that increasingly focus on applying knowledge, justifying decisions, and demonstrating independent clinical reasoning.

Make Students Challenge AI-Generated Recommendations

Nursing educators can also use AI’s limitations as teaching opportunities.

Elliott describes Spot the Hallucination and Spot the Error assignments in which students evaluate AI-generated clinical recommendations containing plausible but potentially dangerous mistakes.

Examples include inappropriate medication combinations, incorrect IV infusion rates, and inaccurate triage priorities.

Students compare the recommendations with approved drug references and evidence-based clinical guidelines to identify potential safety violations.

Elliott also recommends introducing new clinical information during an AI-assisted exercise to see whether the tool adjusts its recommendations appropriately.

For example, students might present a worsening blood pressure or declining urine output and evaluate whether AI recognizes the urgency of the patient’s changing condition.

Murphy similarly teaches students to question AI-generated information, even when the tool has been instructed to rely on course materials.

She encourages students to ask whether the information aligns with their learning, can be verified through reliable references, makes clinical sense, omits important details, or could potentially harm a patient.

Danielle K. Miller, DNP, MSN, RN, founder of The Pivot Nurse and a former adjunct nursing faculty member and clinical nurse educator, recommends giving students an AI-generated response alongside the original research and requiring them to critique it.

Students should determine whether the response accurately represents the findings, overlooks study limitations, or makes recommendations that extend beyond the evidence.

Miller also encourages students to consider whether recommendations are realistic for individual patients.

A treatment plan may appear clinically appropriate but become impractical when patients cannot afford medications, access services, or follow recommended interventions.

These exercises connect AI literacy with evidence-based practice, health equity, and patient-centered decision-making.

Using AI to Strengthen Evidence-Based Practice

Beyond clinical simulations, AI can help students organize research and identify literature that deserves closer examination.

Miller recommends introducing the technology after students have developed a focused clinical question and retrieved relevant research through an academic database.

For example, a student researching postpartum depression could use AI to help organize retrieved articles by patient population, intervention, and study design.

The student would then review the original studies, determine which findings are relevant, and explain why they included or excluded particular articles.

AI may make the research process more manageable, but it cannot replace evaluating study quality or determining whether evidence applies to a clinical situation.

Miller also recommends testing students’ clinical reasoning by changing elements of a patient case.

What happens when a patient develops another condition, new assessment findings emerge, or a recommended resource becomes unavailable?

Assessing Clinical Judgment Without AI

Even when AI supports learning, faculty must determine whether students can apply their knowledge independently.

Murphy uses unfolding case studies, prioritization activities, simulations, classroom discussions, and Next Generation NCLEX-style questions to evaluate students’ understanding.

See also  7 Innovative Ideas to Engage Your Nursing Class with AI

She may ask students to identify the most concerning assessment finding, explain why it takes priority, and determine which action the nurse should take first.

Elliott emphasizes assessments that require real-time decisions without AI assistance, including proctored examinations, clinical skills demonstrations, oral questioning, and immersive simulations.

In high-fidelity simulations, students must recognize changes in patient conditions and respond without stopping to consult an AI tool.

Kolde similarly emphasizes evaluating competence beyond examination performance.

The objective is to prepare graduates who can make safe, effective decisions in clinical practice.

Addressing Accuracy, Bias and Patient Safety

Generative AI can produce inaccurate information, fabricate citations, or recommend interventions that are inappropriate for a particular patient.

Its responses may also reflect biases present in training data, raising concerns about how recommendations could affect patients from different backgrounds.

Elliott emphasizes teaching students to recognize how historical disparities in health care may influence algorithmic recommendations.

Miller recommends incorporating a health equity perspective so students consider whether AI-generated recommendations account for patients’ financial circumstances, access to care, and ability to follow treatment plans.

Patient privacy is another critical concern.

Faculty must establish clear expectations about what information students may enter into AI tools and ensure that activities comply with institutional policies governing patient information and data security.

Academic integrity policies should distinguish between permitted uses, such as brainstorming or developing study materials, and activities requiring independent student work.

Kolde stresses the importance of faculty development, noting that educators need opportunities to understand AI’s capabilities and limitations before integrating it into their teaching.

She also encourages institutions to consider differences in students’ technological experience and seek input from clinical partners and employers when developing AI-related competencies.

How Nursing Faculty Can Get Started

Educators do not need to redesign an entire course to begin exploring AI-supported instruction.

Murphy recommends starting with a focused activity, such as developing practice questions or a clinical case study aligned with existing learning objectives.

Faculty should review generated content for accuracy and relevance before introducing it to students.

Miller suggests asking students to evaluate how accurately AI summarizes a research article, identifying omissions and checking evidence.

Elliott recommends beginning with a low-stakes pilot and selecting technology based on the clinical judgment skill students need to demonstrate.

Kolde also encourages faculty to experiment with AI tools, evaluate their responses and pursue professional development opportunities.

AI can help faculty develop engaging activities, personalize learning, and expose students to new ways of evaluating information. But its educational value depends on whether those activities require students to think critically, verify evidence, and justify their decisions.

For nursing faculty preparing the next generation of NPs and other nurses, the challenge is teaching students not only how to use AI, but also when to question it, reject its recommendations, and trust their own clinical judgment.

Renee Hewitt
Follow
Share