Nursing undergraduates’ experiences of generative AI-assisted learning in the classroom environment: a qualitative study

Abstract

Background

Integrating generative artificial intelligence (GenAI) into course learning has emerged as a significant trend in the future development of nursing education. Students are central to learning activities, and their authentic experiences with GenAI-assisted learning are essential for understanding how educational innovations are perceived and implemented. However, there remains a paucity of qualitative studies focusing on students with GenAI-assisted learning experiences in the classroom environment. This study therefore aimed to explore nursing undergraduates’ experiences, perceptions, and opinions following their participation in structured GenAI-assisted learning in a formal classroom environment involving both instructors and peers.

Methods

Based on the guidance of the Unified Theory of Acceptance and Use of Technology (UTAUT), a semi-structured interview outline was designed. Using purposive sampling, 12 nursing undergraduates with learning experiences assisted by GenAI in the classroom environment were recruited for semi-structured interviews, and the data were analyzed using the six-phase reflexive thematic analysis method.

Results

Analysis yielded four themes: (1) Perceived usefulness and ease of use, (2) Social facilitation and resource support, (3) Risk concerns, and (4) Expectations and suggestions.

Conclusion

Structured GenAI-assisted learning was perceived to support nursing undergraduates’ classroom learning and skill development. Future efforts may focus on optimizing teaching strategies, strengthening ethical guidance, and developing nursing-specific GenAI tools to promote the responsible and effective integration of GenAI into nursing education.

Background

The global healthcare sector is undergoing unprecedented digital transformation, imposing new and heightened demands upon the professional competencies of nursing staff [1]. In the future, nurses will not only need solid clinical knowledge and proficient procedural skills, but also strong information literacy, critical thinking, a capacity for lifelong learning, and effective decision-making skills in order to navigate increasingly complex clinical situations [2–5]. Against this backdrop, higher nursing education faces an urgent imperative to reform pedagogical models and enhance teaching quality, thereby cultivating innovative nursing professionals capable of adapting to and spearheading future healthcare developments [2, 6]. Exploring and integrating emerging technologies to innovate teaching methods has become a focal point within the global nursing education community [7].

In recent years, generative artificial intelligence (GenAI) technologies, exemplified by large language models, have achieved breakthrough progress. Their formidable capabilities in natural language understanding and generation have provided a technological pathway for innovation in nursing education [8]. GenAI has been defined as a category of artificial intelligence technologies that are capable of learning the latent distributions and intricate patterns within massive datasets [9]. By modeling this learned information, these algorithms could generate new, original content across various modalities, such as text, images, audio, and video [9]. Given the multifaceted advantages of GenAI (e.g., possessing powerful content generation and interactive capabilities), its integration into nursing education has become a research focus [10, 11].

Preliminary evidence has confirmed the positive role of GenAI in nursing education. For instance, Fung et al. [12] employed GenAI to simulate patient teaching scenarios, which effectively improved nursing students’ clinical perceptual abilities. Similarly, Song et al. [13] designed a series of five pedagogical activities incorporating GenAI into classroom instruction, leading to significant enhancements in students’ higher-order thinking skills and artificial intelligence (AI) literacy. In the context of clinical training, Arkan et al. [14] utilized ChatGPT to train students on the nursing process, which markedly strengthened their problem-solving and nursing process competencies. Furthermore, Tseng et al. [15] introduced ChatGPT into the teaching of academic report writing, observing a substantial improvement in students’ case report writing proficiency. These studies underscore the considerable promise of GenAI as a transformative tool in nursing education, yielding significant and demonstrable improvements in students’ learning outcomes.

Although existing studies have affirmed the developmental prospects of GenAI in nursing education and explored its application in clinical simulations or extracurricular self-directed learning [16, 17], it must be noted that previous research has largely neglected students’ actual experiences within regulated formal classroom environments, characterized by teacher guidance and peer interaction. Students’ genuine experiences in various contexts are essential for assessing the effectiveness of educational innovations, as they represent the cornerstone of evaluating the impact of teaching advancements [18]. However, current understanding of students’ specific perceptions regarding the use of GenAI in supervised classroom settings remains limited [19, 20].

Furthermore, while GenAI presents educational opportunities, it concurrently introduces significant challenges concerning academic integrity and the accuracy of medical information [21]. Existing qualitative research has predominantly focused on settings outside the classroom, often overlooking the complexity of the classroom environment [17, 22]. There is insufficient understanding of the specific coping mechanisms students employ when facing these risks under the direct influence of instructors and peers [23]. Therefore, it is of particular importance to investigate the specific learning experiences, perceptions, and suggestions of nursing undergraduate students regarding the use of GenAI in supervised classroom environments. Only by gaining an in-depth understanding of the interaction between GenAI and students within the classroom environment can educators develop teaching strategies that both maximize its appropriate use and effectively mitigate risks, thereby promoting the responsible and learner-centered implementation of GenAI-assisted instructional strategies [24].

Building on our previous work that proposed five structured teaching strategies for GenAI-assisted learning (resource expansion, scenario simulation, role-playing, instant question answering, and collaborative content generation) in the classroom environment [13], this study therefore aimed to explore nursing undergraduates’ experiences, perceptions, and opinions following their participation in such strategies in a formal classroom environment involving both instructors and peers.

Theoretical framework

The Unified Theory of Acceptance and Use of Technology (UTAUT) was proposed by Venkatesh et al. [25]. The UTAUT comprises four core constructs: performance expectancy, effort expectancy, social influence, and facilitating conditions. Performance expectancy refers to the degree to which an individual perceives that using a particular technology will help them to enhance their performance in work-related or other activities. Effort expectancy refers to the degree of effort an individual perceives is required to use a particular technology. Social influence refers to the perceived pressure an individual feels from significant others to use the technology. Facilitating conditions refers to the assessment by an individual of whether the organization and technical facilities they are in are adequate to support their use of that technology to the extent required. It has been widely applied in research across numerous domains to explain individuals’ acceptance behavior toward various technologies [26, 27]. The UTAUT theory provides a systematic theoretical framework for exploring nursing undergraduates’ perceptions of GenAI use in the classroom environment.

Methods

Design

This study used a descriptive qualitative design in accordance with the Standards for Reporting Qualitative Research (SRQR) [28]. Semi-structured interviews were undertaken with nursing undergraduates who had experience of GenAI-assisted learning. This methodology was selected to comprehensively explore nursing undergraduates’ experiences, perceptions, and opinions regarding their use of GenAI-assisted learning in the classroom environment.

Setting and participants

This study was conducted at the school of nursing of a medical university in Anhui Province, China. From September to December 2024, we conducted a one-semester, GenAI-assisted classroom experiment within the course of Thinking and Communication, a national first-class undergraduate course, for 66 third-year nursing undergraduates. The GenAI integration followed the five structured teaching strategies described in the Background section (resource expansion, scenario simulation, role-playing, instant question answering, and collaborative content generation). To gain an in-depth understanding of these students’ experiences, perspectives, and opinions regarding the curricular integration of GenAI-assisted learning, we employed a purposive, maximum-variation sampling strategy guided by the principle of data saturation among the 66 participants who had completed the course [29]. Twelve students were ultimately interviewed to capture the widest possible spectrum of experiences associated with GenAI-assisted learning in the classroom environment.

Data collection

Data were collected through semi-structured interviews conducted by two research assistants trained in qualitative research methods. The study was approved by the Ethics Committee of the medical university (Approval No. 82240172). The semi-structured interview guide was developed based on the four core constructs of the UTAUT. The questions in the interview guide were designed as open-ended to encourage participants to provide detailed, narrative responses. Furthermore, we added two questions in the guide to: (1) explore the potential risks and limitations associated with the use of GenAI in the classroom, and (2) assess learning satisfaction and suggestions concerning the future application of GenAI in nursing education. After drafting the semi-structured interview guide, two research assistants trained in qualitative research conducted pilot interviews with two eligible nursing undergraduates. Although the pilot findings were not included in the formal analysis, they were used to refine the interview guide and finalize its content. The content of the interview guide is provided in Table 1.

Table 1.

The interview guide developed based on the UTAUT

Construct in the UTAUT	Aim of interview	Question in the interview guide
Social influence	To explore how external social factors (such as recommendations from teachers, peers, or senior students) influence students’ adoption and use of GenAI.	How did you initially learn about or begin using GenAI applications in your study? Did anyone encourage you to use it?
Effort expectancy	To explore the students’ perceived ease of use and user-friendliness of GenAI tools.	Could you please describe your experience when using the GenAI application in the classroom?
Performance expectancy	To explore the students’ beliefs regarding whether GenAI could enhance their learning efficiency, academic performance, and problem-solving abilities.	In what ways do you believe using GenAI in the classroom has helped your learning?
Facilitating conditions	To assess the roles of external resources (e.g. technical support, training, and access to necessary equipment or software) in facilitating the students’ use of the GenAI in classroom learning.	What support or resources did you receive to help you use the GenAI applications in your study in the classroom?
Additional (beyond UTAUT)	To explore potential risks and limitations of using GenAI in the classroom.	What concerns do you have about using GenAI applications in classroom learning?
Additional (beyond UTAUT)	Assess learning satisfaction and gather suggestions for its future use in nursing education.	What suggestions or expectations do you have for the future application of GenAI in nursing education?

Note: UTAUT, Unified Theory of Acceptance and Use of Technology. The final two questions (beyond UTAUT) capture potential risks and future perspectives, which are not explicitly covered by the original framework but are essential for understanding students’ GenAI-assisted learning experiences

Recruitment and data collection during the interview period ran from February to March 2025, immediately following the course examination. All interviews were conducted in a private, quiet room. Prior to each interview, the research assistants explained the study purpose and procedures and obtained written informed consent. Participants were assured that their responses would not affect their course grades.

Data saturation was assessed during the concurrent process of data collection and initial thematic coding. After each interview, the two research assistants independently reviewed the interview notes and preliminary codes to monitor the emergence of new themes. Saturation was determined to have been reached after the eleventh interview, as no new themes or codes were identified. To confirm this, a twelfth interview was conducted; it yielded no new information, confirming that saturation had been achieved. Consequently, the final sample size for this study comprised 12 students (coded S1-S12), including 9 female and 3 male participants (see Table 2). Each interview lasted approximately 25 to 35 min. With explicit participant consent, all interviews were audio-recorded. Subsequently, these recordings were encrypted and securely stored on a password-protected computer to safeguard participant privacy and data security.

Table 2.

The characteristics of participants

No.	Gender	Age	Position in class
S1	Female	22	None
S2	Male	21	None
S3	Male	21	None
S4	Female	20	Class monitor
S5	Female	21	None
S6	Female	21	None
S7	Female	21	None
S8	Female	21	Class representative for learning
S9	Male	21	None
S10	Female	22	None
S11	Female	21	None
S12	Female	21	None

Data analysis

Data were analyzed following the six-phase reflexive thematic analysis procedure outlined by Braun and Clarke [30], an approach widely endorsed in healthcare and nursing qualitative inquiry [31, 32]. All recorded interviews were transcribed verbatim within 24 h of completion, anonymized, and imported into NVivo 20 for systematic storage and coding. Two research assistants independently conducted open coding line by line; discrepancies were resolved through discussion, consulting a third researcher where necessary. Subsequently, codes were clustered into candidate sub-themes based on conceptual similarity and recurring patterns. These sub-themes underwent iterative review and consolidation into overarching themes encompassing the core phenomena of the study.

Rigor

To ensure the rigor of the research, two interviewers received systematic qualitative interview training and were supervised throughout by a Master’s supervisor with extensive experience in qualitative research. The interview guide was finalized by the research team based on the UTAUT, an extensive literature review and multiple rounds of group discussions.

Reflexivity

The interviewers (ZW, RZ) were postgraduate nursing students from the same institution as the participants, with no prior evaluative relationship. The corresponding author (LZ) was the course instructor, creating potential social desirability bias.

To mitigate bias: (1) interviewers assured participants of grade independence and anonymity; (2) leading questions were avoided, and critical feedback was actively encouraged; (3) initial coding was performed independently by DS and YZ (not involved in data collection), with LZ excluded from initial coding; (4) an external qualitative researcher reviewed preliminary findings; (5) member checking with three participants confirmed resonance.

We recognize our professional investment in GenAI may have introduced affirmative bias. We actively sought disconfirming evidence, such as students’ risk concerns, as reflected in the Results section. Additionally, our interpretation of “teacher guidance” may have been influenced by the specific cultural context of China. Readers are encouraged to critically evaluate this perspective in light of their own educational backgrounds.

Results

The analysis generated four themes and thirteen sub-themes: (1) Perceived usefulness and ease of use (increased classroom engagement, enhanced learning efficiency, strengthened communication skills, improved thinking abilities, and straightforward operation), (2) Social facilitation and resource support (teacher recommendations and guidance, group effect and peer influence, accessibility and zero cost), (3) Risk concerns (content accuracy risks, dependency risks, and data privacy breaches), and (4) Expectations and suggestions (strengthening professional relevance and optimization of teaching strategies). The mapping between the UTAUT constructs and the themes identified through the analysis is presented in Table 3.

Table 3.

Mapping of UTAUT constructs to the identified themes

Construct in the UTAUT	Corresponding theme(s) in findings
Performance expectancy	Theme 1: Perceived usefulness and ease of use (sub-themes: increased classroom engagement, enhanced learning efficiency, strengthened communication skills, and improved thinking abilities)
Effort expectancy	Theme 1: Perceived usefulness and ease of use (sub-theme: straightforward operation)
Social influence	Theme 2: Social facilitation and resource support (sub-themes: teacher recommendations and guidance, group effect and peer influence)
Facilitating conditions	Theme 2: Social facilitation and resource support (sub-themes: accessibility and zero cost)
Additional (beyond UTAUT)	Theme 3: Risk concerns (content accuracy risks, dependency risks, and data privacy breaches)
Additional (beyond UTAUT)	Theme 4: Expectations and suggestions (strengthening professional relevance and optimizing teaching strategies)

Perceived usefulness and ease of use

Increased classroom engagement

As an emerging technology in classroom teaching scenarios, GenAI stimulated students’ curiosity. Students described GenAI as making classroom learning more interactive and enjoyable compared to traditional lecture-based teaching. The teacher integrated GenAI-assisted activities such as simulated patient dialogues and instant question answering, which provided immediate, personalized responses. This novelty and responsiveness reduced passivity and encouraged even quiet students to participate. S2: “The teacher set up many teaching activities related to GenAI, which I found very interesting. The classroom environment is no longer so dull; everyone is no longer simply listening to the teacher but actively participating in interactions. Even those students who usually don’t like to speak up will try to ask questions or share their thoughts through AI.” S5: “This was the first time I saw a teacher use GenAI in the classroom. I found it quite novel, and I look forward to the surprises AI will bring in every class…”.

Enhanced learning efficiency

Most students reported that GenAI could retrieve vast amounts of information, provide real-time answers, and help students understand complex concepts. Participants perceived that GenAI shortened the information retrieval process, simplified learning tasks, and improved their learning efficiency by providing synthesized answers and structured explanations. S1: “Previously, it took me several days to complete the course summary independently. After using GenAI, I finished the course summary in just one day, which greatly improved my learning efficiency.” S4: “GenAI has a strong ability to collect information, saving me a lot of time, so I have more time to do other things.” S8: “Sometimes I find the questions raised by the teacher a bit obscure and difficult to understand. I use AI to help me break down and interpret the teacher’s meaning. With its help, I can complete a difficult problem relatively smoothly.”

Strengthened communication skills

Some students perceived that using GenAI helped strengthen their communication skills. The tool could simulate patient interactions, allowing students to rehearse dialogues in a low-stakes environment. Moreover, formulating effective prompts and engaging in iterative questioning trained their ability to express needs clearly and ask follow-up questions. S5: “GenAI can simulate patients to conduct mock dialogues with me, and this process helps to exercise my communication skills.” S2: “Sometimes, regarding a specific issue, I will ask and follow up with multiple rounds of questions to the AI so that it can better understand my meaning. I feel my questioning techniques and expression abilities have been trained.”

Improved thinking abilities

Some students emphasized that their thinking had improved significantly compared to before. They observed that GenAI’s well-structured and logical answers helped them adopt more systematic reasoning; exposure to up-to-date research and diverse perspectives also sparked innovative ideas. S6: “AI’s answers are always very comprehensive and highly logical. Now, when I think about problems, I also focus on systematic thinking and logic.” S11: “When it answers questions, it always provides some of the latest research or cutting-edge information, which sometimes sparks some of my innovative ideas.”

Straightforward operation

Students generally described GenAI as having a concise interface, clearly organized functional modules, and a low technical threshold. They perceived that they did not need to master professional programming skills or complex operation procedures; efficient human-computer interaction can be achieved merely through simple interaction methods such as natural language input and voice commands. S1: “The teacher spent only 5 minutes introducing the basic operations in class, and we were able to independently use AI for case analysis and knowledge point queries.” S7: “As long as you type out the question you want to ask or say it by voice, it can quickly give a clear and organized answer. The interface does not have superfluous complex buttons; even classmates who are not very good at using electronic products can get started within a short time.”

Social facilitation and resource support

Teacher recommendations and guidance

Some students reported that the teacher’s recommendation and professional guidance were key support factors that encouraged them to begin using GenAI effectively. Students expressed that they might not have attempted to use GenAI for learning if the teacher had not explicitly introduced it and demonstrated its applications. S4: “My first contact with GenAI was recommended by the teacher in class; they also explained the specific usage methods and scenarios, which gave me the desire to try using it.” S2: “If the teacher hadn’t specifically taught us how to use it to help learning in class, I probably wouldn’t have used it for studying. It was after the teacher taught us some techniques that I truly felt it was useful for learning.”

Group effect and peer influence

Some students reflected that peer sharing and group feedback created a positive group effect. Such group-level interaction lowered the threshold for adopting new technologies and improved students’ confidence in using GenAI through the exchange of experiences. Peer sharing and group discussions created a positive feedback loop. S10: “The first time I used this tool was when I saw my classmate using it. I thought it was quite convenient, so I thought I’d try it myself.” S9: “During group discussions organized by the teacher in class, everyone will actively share AI usage skills, such as how to input precise questions and how to filter AI-generated content. This made me avoid many detours, and my ability to use AI tools has also been greatly improved through communication.”

Accessibility and zero cost

Students perceived that convenient access to GenAI tools and the lower economic cost reduced the usage threshold of the technology. Students reported that they could access these tools via web browsers or mobile apps to carry out learning activities anytime and anywhere, which further strengthened their acceptance of the technology and frequency of use. S5: “The AI tool I often use can be opened directly in the browser without downloading specific software. No matter where I am, I can log in to query relevant knowledge points; it is particularly convenient.” S8: “Currently, the basic functions and installation channels of several popular AI tools are free. This free access means I have no reason to refuse using it.”

Risk concerns

Content accuracy risks

Students were aware that GenAI could generate incorrect or fabricated information, such as non-existent references or inaccurate details. This risk prompted them to cross-check GenAI outputs with reliable sources. S12: “Sometimes the references it provides are incorrect, and I simply cannot find the papers it lists.” S9: “It is very fast to use for checking information, but I dare not use it directly. I am always worried that it will fabricate non-existent literature or data, so when I encounter uncertainties, I will verify the accuracy of the answers through multiple sources.”

Dependency risks

Some students were concerned that long-term use of GenAI could lead to dependency. They noticed that their inclination to solve problems independently had weakened after frequent use. Some described feeling “lost” or inefficient without GenAI assistance. This dependency risk was recognized even by students who acknowledged GenAI’s benefits. S1: “Although using GenAI saves time, I feel that I am becoming increasingly reliant on AI. Once, without its assistance, sometimes I don’t know how to carry out learning efficiently.” S5: “Although I am now consciously controlling myself not to overuse AI, AI is really convenient.”

Data privacy breaches

In the process of interacting with GenAI, students inevitably input a large amount of highly personalized and academic information, and they generally worry whether this private information will be obtained by third parties without authorization. Some deliberately avoided sharing sensitive ideas. This privacy concern limited the depth of their interaction with GenAI. S4: “I often use GenAI tools to complete homework and find information. Every time the thinking process and modification records I input are saved by the tool, I am a bit worried that this data might be leaked to third parties.” S11: “I always feel that every sentence I say to it is recorded, so regarding some key core ideas, I dare not input them; I feel it is unsafe.”

Expectations and suggestions

Strengthening professional relevance

Some students described GenAI-generated content as lacking specificity and practical relevance for nursing, making it difficult to fully meet their learning and practice needs. They wished for GenAI tools tailored to nursing and clinical contexts. They envisioned GenAI that could directly output nursing-standard procedures, care plans, or patient education materials aligned with industry guidelines. S3: “I hope AI tools can specifically optimize their knowledge base for the nursing profession. For example, when querying the ‘operating room nursing safety verification process’, it can directly output operating steps that conform to nursing industry standards, rather than general medical advice.” S11: “Current GenAI tools are too universal; it would be better if there were a GenAI tool developed specifically for medical students or nursing students.”

Optimizing teaching strategies

Students expected GenAI to be promoted to other courses and integrated more deeply into the design of nursing teaching activities. They called for tasks that require deeper human-AI collaboration, such as complex clinical simulations or peer-AI group work. S4: “I hope teachers can design more learning tasks that require us to think actively and interact deeply with GenAI and classmates.” S6: “I look forward to AI entering our learning process in more forms in the future.” S10: “I hope to increase the design of clinically relevant learning activities with high fidelity and strong practical relevance, helping us to adapt to the complexity and uncertainty of clinical nursing work in advance.”

Discussion

By understanding nursing undergraduates’ experiences with structured GenAI-assisted classroom learning, this study provides preliminary, context-specific insights into the integration of GenAI in the classroom environment of nursing education. The findings showed that participants generally held positive attitudes toward structured GenAI-assisted learning in the classroom environment, with perceived usefulness and ease of use serving as the core drivers for their use of GenAI, while teachers’ guidance and social support constituted crucial external environmental factors. Nevertheless, participants raised concerns about content accuracy, dependency and data privacy breaches. They also offered suggestions for optimizing teaching strategies and developing nursing-specific GenAI applications.

Perceived usefulness and ease of use

As presented in Theme 1, students perceived GenAI as a valuable and easy-to-use tool that enhanced their learning experience. This finding directly aligns with the UTAUT constructs of performance expectancy (the belief that using GenAI improves learning outcomes) and effort expectancy (the perceived ease of using the technology). Specifically, participants reported that using GenAI in the classroom helped improve learning efficiency and classroom engagement, which is consistent with the findings of Carmi [33]. Nursing curricula often encompass a substantial amount of basic medical knowledge and complex nursing procedures, resulting in a heavy information load [34]. As an intelligent assistive tool, GenAI can swiftly retrieve and integrate vast quantities of information, helping students shorten the time spent on data searching and knowledge organization, thereby enabling them to devote more energy to their studies [35]. Positive emotional experiences serve as the foundation for maintaining long-term learning engagement. GenAI-assisted teaching enhances classroom immersion and interest through dynamic interaction and personalized feedback, effectively stimulating students’ learning motivation [36].

Moreover, participants indicated that the process of interacting with GenAI contributed to the development of communication skills and thinking abilities. This finding aligns with positive perspectives of Boscardin et al. [37], yet appears to contrast with the conclusions of Saglam and Kalanlar [38], who found that nursing students reported weakened critical thinking and over-dependence on ready-made information. This divergence underscores the pivotal role of pedagogical context: In our study, GenAI was integrated into structured classroom activities that included teacher guidance, peer collaboration, and iterative inquiry tasks. These activities encouraged students to question, verify, and synthesize AI-generated content rather than simply accept it. In contrast, when students use GenAI independently without sufficient guidance, they may be more likely to rely passively on AI-generated responses, which could contribute to the decline in thinking abilities reported by Saglam and Kalanlar [38]. Hence, within the bounds of this study, the effect of GenAI on nursing students’ thinking abilities is not solely inherent to the technology itself but may partly depend on how it is integrated into the learning environment, the nature of student-AI interactions, and the presence of structured educator guidance. In addition, some students expressed that GenAI required minimal technical proficiency in this study. This view was consistent with Salama et al. [39], who observed that the low technical barrier of GenAI may reduce the apprehension some nursing students feel toward high-tech tools. The low-threshold, low-cognitive-load experience may promote the acceptance of GenAI among nursing students [40].

Social facilitation and resource support

Aligned with Theme 2, teacher guidance and peer influence were identified as significant external factors influencing students’ acceptance and use of GenAI. These findings reinforce the UTAUT, highlighting the role of social influence (encouragement from teachers and peers) and facilitating conditions (access to resources and support). Participants emphasized that teacher guidance and group effects were significant factors influencing individual use of new technologies, which is consistent with findings from Arkan et al. [14], whose study showed that teacher-guided intervention increased students’ positive attitudes toward GenAI while their negative attitudes decreased. This study emphasized teachers’ guidance in the classroom environment and the interactive use of GenAI by students, a process that required students to participate actively, ask questions, and integrate information. In contrast, students in the study by Summers et al. [22] expressed concerns regarding the substitutive and passive use of GenAI without teacher guidance. Therefore, within the context of this study, the impact of GenAI in nursing education ultimately depends on how nurse educators teach students to utilize it.

Indeed, when teachers integrated GenAI into instructional design and provided professional guidance, they not only conveyed the practical and educational value of the technology to students but also helped them master correct usage methods, thereby avoiding blind exploration [41]. Experience sharing and skill exchange within groups lowered the psychological threshold for individuals to trial new technologies and fostered the enhancement of collective efficacy [42]. Furthermore, the accessibility of tools and their cost-free characteristics also served as important drivers for individual use of new technologies; if access to technological tools was cumbersome or costs were high, the willingness to use them was severely hindered [41].

Risk concerns

As reported in Theme 3, participants expressed clear concerns about content accuracy, dependency, and data privacy, which align with Stratton-Maher and Kelly [43]. Although not originally part of the UTAUT, these risk concerns emerged as a salient theme in our data, reflecting specific challenges posed by GenAI in nursing education and indicating potential barriers to continued use and trust. The nursing profession demands high standards of knowledge, accuracy, and rigor; consequently, the risk of content inaccuracy associated with GenAI could pose risks for future clinical practice [44]. While participants perceived GenAI as enhancing learning efficiency, it could also lead some students to form dependencies, thereby weakening their capacities for independent thinking, information screening, and problem-solving [45]. Furthermore, if there were risks of leakage when students used GenAI tools, this would not only infringe upon student rights but also reduce their enthusiasm for use, becoming an obstacle to the deep integration of the technology [46]. Hence, nursing educators should establish robust content verification mechanisms, train students in critical thinking, clarify the boundaries of GenAI-assisted teaching and learning, and strengthen education on data privacy to ensure responsible use.

Expectations and suggestions

As shown in Theme 4, participants generally reported that current GenAI models struggled to provide precise guidance aligned with nursing industry standards, and they called for more varied integration strategies. This finding aligns with previous studies [47, 48]. These expectations and suggestions go beyond the UTAUT and may offer valuable guidance for pedagogical innovation and tool development in nursing education. Therefore, nurse educators should lead interdisciplinary collaboration, working alongside AI developers and clinical experts to co-develop nursing-specific AI models focused on the core needs of the nursing profession. Furthermore, they should actively explore modes of integration and innovate teaching activities to foster the deep integration of GenAI with curriculum instruction [49].

Strengths and limitations

To the best of our knowledge, this study is one of the few exploratory studies conducted among nursing undergraduates’ experiences with GenAI-assisted learning in the classroom environment. In this study, an interview guide was designed based on the core constructs of the UTAUT. The findings provided a structured understanding of nursing undergraduates’ experiences, perspectives, and opinions following their participation in GenAI-assisted learning, and offered meaningful guidance for the future application of GenAI in nursing education.

However, there were several limitations in this study. Firstly, this study was conducted at a medical university, which may limit the transferability of the findings. Secondly, while teacher guidance was identified as a core finding in this study, the perspectives of teachers were not included, which could constitute a substantive omission. Future research should therefore incorporate paired teacher-student accounts to broaden the research perspectives, and expand the sample size and conduct multicenter studies to test transferability.

Conclusion

This qualitative study found that nursing undergraduates generally held positive attitudes toward structured GenAI-assisted learning in the classroom environment. Participants perceived GenAI-assisted learning as supporting classroom engagement, learning efficiency, and the development of communication and thinking skills. However, they also expressed concerns and offered suggestions regarding the future use of GenAI in nursing education. Moving forward, nursing educators and institutions should continue to optimize GenAI-assisted teaching strategies, provide clear guidance on responsible use, and support the development of nursing-specific GenAI tools that are aligned with educational needs and professional standards.

Abbreviations

AI

Artificial intelligence

GenAI

Generative artificial intelligence

GenAI-assisted

Generative artificial intelligence-assisted

SRQR

Standards for reporting qualitative research

UTAUT

Unified theory of acceptance and use of technology

Author contributions

DS & YZ: Conceptualization, Methodology, Formal analysis, Writing-original draft. XX & ML: Data curation, Investigation, Project administration. ZW & RZ: Data curation, Investigation, Visualization. PZ: Project administration. LZ: Conceptualization, Methodology, Resources, Project administration, Writing-review & editing, Supervision.

Funding

This study was supported by the Ministry of Education, P. R. China (National First-Class Online Course Program) (Certificate No. 2020110370), Anhui Provincial Department of Education (Teaching Quality Project) (No. 2023jyxm0232; 2024sswjc065; 2024jcjs052), Anhui Provincial Department of Education (New Era Education Quality Project) (No. 2024szsfkc047) and the Research Program of School of Nursing, Anhui Medical University (No. Hlqm12025014; Hlqm12025023; Hlqm22025004; Hlqm22025017).

Data availability

Datasets are available upon reasonable request by contacting the corresponding author.

Declarations

Ethics approval and consent to participate

The protocol of this study has been approved by the Ethics Committee of Anhui Medical University (Approval number: 82240172) in accordance with the Declaration of Helsinki. All participants were provided with comprehensive information about the study objectives, methodology, potential risks, and the measures to ensure confidentiality. Written informed consent to participate was obtained from all participants before the initiation of the study.

Consent for publication

All participants provided written informed consent for study participation and the publication of anonymized data.

Competing interests

The authors declare no competing interests.