Artificial intelligence in nursing practice: a qualitative study of nurses’ perspectives on opportunities, challenges, and ethical implications

Abstract

Background

The study aims to explore nurses’ views on the effects of artificial intelligence (AI) in nursing, focusing on their understanding, practical applications, ethical considerations, and perceived opportunities and threats.

Methods

This qualitative study used semistructured interviews to gain comprehensive insights from clinical nurses, adhering to the Standards for Reporting Qualitative Research for methodological rigor. After obtaining ethical approval, researchers conducted semistructured interviews with 25 clinical nurses. The interviews explored nurses’ perceptions of AI, including its basic concepts, applications in nursing practice, ethical and social implications, and potential benefits and drawbacks.

Results

The analysis identified four overarching themes: (1) Nurses’ Conceptualizations of Artificial Intelligence, (2) Opportunities of AI in Nursing Practice, (3) Threats of AI in Nursing Practice, and (4) Ethical and Psychological Concerns in AI-Based Nursing Practice. The findings revealed that nurses had a foundational understanding of AI and its definitions. They acknowledged both the positive and negative impacts of AI technologies on their practice. Nurses expressed that AI could reduce workload, enhance patient care, and improve efficiency. However, they also articulated significant threats, including concerns over professional redundancy, emotional disconnection in caregiving, de-skilling, and the risk of dehumanizing the healthcare environment. Additionally, ethical and psychological concerns emerged, such as ambiguity in accountability, threats to data security and patient safety, unsuitability in psychiatric care contexts, staff surveillance anxiety, and risks of misuse or systemic bias.

Conclusion

The study concluded that while nurses possess a basic understanding of AI, the effective and ethical integration of AI technologies in nursing requires targeted training, institutional preparedness, and robust interdisciplinary collaboration. To ensure AI complements rather than compromises nursing values, it is imperative to equip nurses with skills in digital literacy, ethical reasoning, and critical engagement with AI tools. The findings highlight the necessity of structured education programs and policy development that address both the technological and humanistic dimensions of AI use in healthcare. Future research should actively incorporate patient and public voices to ensure that AI-driven transformations in care remain aligned with the principles of patient-centeredness and human dignity.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12912-025-03775-6.

Tweetable abstract

#Artificial Intelligence in Nursing Practice: A Qualitative Study of Nurses’ Perspectives on Opportunities, Challenges, and #Ethical Implications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12912-025-03775-6.

Introduction

Artificial intelligence (AI) first gained prominence with Alan Turing’s machine intelligence test in 1950, impacting a wide range of fields, including engineering, biology, neuroscience, sociology, nursing, education, and medicine [1–4]. In healthcare, AI is significantly enhancing clinical decision-making, early diagnosis, medical imaging, drug development, treatment, and health record management [5–10]. This rapid adoption has catalyzed increased research on AI’s impact on health disciplines [11]. Nurses, who are integral to healthcare delivery, are also experiencing this technological shift [12]. AI promises to drive healthcare innovation, enabling individualized, evidence-based care and significantly influencing nursing practice [13, 14]. In the nursing profession, AI systems are increasingly being introduced to support clinical documentation, patient monitoring, and task delegation. However, despite these potential benefits, nursing has often been overlooked in the development and design of AI technologies [15]. While these technologies promise to reduce workload and enhance care quality, they also raise critical questions about autonomy, accountability, and the human aspects of nursing care. Despite the expanding use of AI in clinical settings, there is a notable lack of empirical research reflecting nurses’ perspectives, particularly in the context of everyday practice. This study addresses this gap by exploring how nurses perceive the opportunities, challenges, and ethical implications of AI integration in nursing practice. It is crucial for nurses to actively participate in the development and application of these technologies [16]. Indeed, many nurses are already utilizing AI to enhance care plans and identify patient risks in clinical settings [17].

Globally, countries are training young nurses to integrate digital technologies into nursing care. Nonetheless, most AI-related nursing research is concentrated in the United States [17]. Buchanan et al. (2020) highlighted the necessity for a shared understanding among nurse researchers and experts to overcome barriers to AI adoption in nursing [5]. Addressing these challenges is vital to unlocking AI’s full potential in clinical practice [18]. This recognition aligns with the current literature, which indicates a notable lack of qualitative research focusing on nurses’ perspectives regarding the integration of AI in nursing. While substantial research exists on AI’s technical applications and its potential across various medical fields, specific insights and expectations of nurses have not been thoroughly investigated. This gap is critical because understanding the unique needs and concerns of nurses is essential for the effective implementation of AI in nursing practice. The implementation of artificial intelligence (AI) in nursing is often positioned as a solution to operational inefficiencies and data-driven decision-making. However, a comprehensive rationale must also consider the deeply relational and affective dimensions of nursing work. Emotional labor—defined as the process by which nurses manage their own emotions to fulfil the emotional needs of patients—is a fundamental component of caregiving that cannot be easily automated or replicated by machines. As Diogo et al. (2021) argue, nurses rely heavily on emotional competence to build trust, convey empathy, and maintain therapeutic relationships, particularly in high-stress or sensitive care environments [19]. Integrating AI into nursing should therefore not aim to replace emotional labor but rather to augment it—by relieving nurses of repetitive physical or administrative tasks, AI could enable more time and cognitive space for relational care. When designed with human-centered values, AI has the potential to support the affective domain of nursing rather than suppress it, creating synergy between technological efficiency and emotional presence.

Therefore, this qualitative study aims to fill this gap by providing an in-depth exploration of nurses’ opinions, views, and expectations regarding AI technologies in nursing and nursing practice. By capturing the voices of nurses, this research could offer valuable insights into how AI can be better integrated into nursing practices, addressing both the opportunities and challenges they perceive.

Background

AI and large language models, such as ChatGPT, hold significant potential to enhance nursing education by offering guidance and resources. There is a potential risk of overreliance on AI-generated data, which can undermine the irreplaceable value of human judgment and expertise. To mitigate these issues, “smart prompt learning” has been proposed, integrating the humanities into AI education to enhance decision-making and critical thinking [4, 20].

In nursing, AI is making significant strides through various innovative applications. For instance, ML methods have proven effective in identifying critical data elements in nursing documentation, helping to detect which patient features are most important in public health applications and allowing for more focused preventive interventions [21]. Additionally, the development of tele-nursing mobile manipulators, such as those presented by Li et al. (2017), can aid nursing care tasks in hazardous clinical environments and potentially automate some tasks, thereby improving safety and efficiency [22].

AI also plays a crucial role in managing hospital-acquired infections. Park et al. (2020) demonstrated that knowledge discovery with ML could facilitate safe catheter removal and urinary tract infection treatment through automated notifications [23]. Furthermore, the early detection of patients who may experience complicated discharges after surgery, as demonstrated by Safavi et al. (2019), can help focus nursing care on those who need it most, thereby enhancing patient outcomes [24]. Predicting and preventing inpatient falls is another area where AI has shown promise. Wang et al. (2019) highlighted that integrative ML prediction could more accurately predict fall risk without manual calculation and provide automatic warning systems, thus improving patient safety [25]. These advancements underline the transformative potential of AI in nursing, enhancing care delivery and improving patient outcomes across various clinical settings.

A scoping review highlighted AI’s potential to improve clinical nursing care, encompassing documentation, nursing diagnoses, care plans, patient monitoring, and care prediction [26]. However, more randomized controlled trials in real-life healthcare settings are needed to validate these findings. Additionally, a bibliometric analysis revealed that most AI-related nursing research is conducted in the United States, with significant contributions from Europe. The study primarily focuses on AI-assisted medical recording and decision-making, indicating a growing interest in AI’s applications in nursing [27].

AI’s role in enhancing nursing education is also significant, with research focusing on profiling and prediction using intelligent agents [28]. The quantitative approach dominates AI-supported nursing research, with health and medicine being the most relevant study areas. Future research should address ethical, methodological, and contextual considerations, emphasizing interdisciplinary approaches to AI applications in nursing [29]. Balancing technological advancements with human judgment is essential as AI continues to evolve. Studies have shown that AI can enhance clinical decision-making but should not replace the invaluable role of human expertise [30, 31].

Comprehensive training programs are necessary to equip nursing professionals with the essential skills and knowledge required for successful AI integration in nursing practice [32]. In educational contexts, Zgambo et al. (2025) revealed that while nursing students increasingly utilize AI for academic tasks, ethical ambiguity and concerns about academic integrity persist, highlighting the need for explicit institutional guidance and reflective digital literacy training. These findings reinforce the broader need for structured AI education among nursing professionals across both academic and clinical settings [33]. Additionally, nurse leaders should support digital health research by providing dedicated time, funding, and professional development opportunities, particularly in underexplored areas [4]. Understanding the perspectives of nursing professionals regarding AI integration is crucial. A study by Rony et al. (2024) explored these viewpoints, highlighting enthusiasm for AI’s potential benefits while emphasizing the importance of ethical and compassionate nursing practice [32]. Fazakarley et al. (2024) also highlighted the need for stakeholder communication to address concerns, mitigate potential risks, and maximize benefits for patients and clinicians alike [34].

The study

This study aimed to explore nurses’ perspectives on the integration of AI technologies in nursing practice. The specific objectives of the study were to:

• Investigate nurses’ opinions and views on the implementation of AI technologies in clinical nursing practice.

• Uncover the perceived benefits and challenges associated with AI integration from the nurses’ standpoint.

• Identify the expectations and requirements of nurses for the practical and ethical use of AI in future nursing practice.

Methods

Design

A qualitative descriptive research design was employed to explore nurses’ perspectives on the integration of AI technologies in nursing practice. This design facilitates a comprehensive exploration of the complex phenomena surrounding AI integration, where a nuanced understanding is essential. The qualitative descriptive research design allowed the researchers to examine the multifaceted nature of AI implementation in nursing by investigating nurses’ opinions, perceived benefits and challenges, and their expectations for the practical and ethical use of AI. The goal was to provide a rich and unembellished portrayal of nurses’ experiences, thereby capturing the complexity inherent in integrating AI technologies into nursing practice. This study adhered to the Standards for Reporting Qualitative Research (SRQR) [35].

Study setting and participants

The study was conducted in various hospital settings in Istanbul, Turkey. Purposive sampling was applied, selecting participants based on the study’s aim. The inclusion criteria were: (1) holding a bachelor’s degree, (2) working at a hospital, (3) residing in Istanbul, and (4) working in different units within the hospital. The study aimed to achieve a diverse representation of nurses from various specialties and departments to gather a comprehensive understanding of their perspectives on AI integration.

Participants were recruited through hospital networks and professional nursing associations. Flyers and email invitations were distributed, providing contact information for the first researcher. Interested participants were assessed for eligibility and scheduled for interviews. Data saturation was achieved after conducting interviews with 25 participants, ensuring a thorough exploration of the topic. The selected participants provided a broad range of insights due to their varied experiences and roles within the hospital settings.

Data collection

Semi-structured and structured interviews were employed to collect data for this study. The researchers developed the interview form based on the existing literature [36–39]. The first part of the form included sociodemographic and professional questions such as age, gender, marital status, education level, and professional experience (overall, hospital-specific, and unit-specific). It also contained 18 questions designed to assess whether nurses had an AI lens before, how they first obtained information regarding AI, if they had received any training on AI, whether they perceived AI as a threat to the nursing profession, and if they supported the integration of AI in nursing.

The second part consisted of open-ended questions designed to elicit more detailed responses:

• How can AI be used in nursing practice? Will it affect the nursing profession? What nursing practices might it involve?

• What do you think about the ethical and social aspects of AI?

• What is your prediction regarding the importance of AI?

Interviews were conducted outside of working hours, with an average duration of 30 min (range: 3060 minutes). Before the start of each interview, participants were encouraged to express their thoughts freely, assured that their names would remain confidential and that their responses would not affect their professional positions or duties. The primary researcher conducted interviews between October and December 2021, also noting participants’ reflections during the interviews.

Ethics approval and consent to participate

Ethical approval

was obtained from the Istanbul University-Cerrahpasa, Social and Human Sciences Ethical Committee (date: 4 October 2021, code: 2021/204). The study was conducted in accordance with the principles of the Declaration of Helsinki. Before the interviews, participants were provided with an information sheet detailing the study objectives and procedures. Written and verbal informed consent was obtained prior to recording, and participants confirmed their willingness to join the study by selecting the “I agree to participate” option at the beginning of the interview. They were informed of their right to refuse any question or to withdraw from the study at any time, in line with the principles of informed consent and respect for autonomy. Confidentiality was ensured, and all personal information was kept strictly anonymous. Both the participants and the researcher retained copies of the signed consent forms to ensure transparency and compliance with ethical standards.

Data management and analysis

The interview transcripts were anonymized and subsequently imported into MaxQDA software for qualitative content analysis. The data analysis process employed an inductive approach to thematic analysis [40]. Initially, two researchers manually decoded and transferred the interviews. They reviewed the interview notes and read the transcriptions carefully to gain a comprehensive understanding of the experiences shared by the participants.

The next steps involved both descriptive and evaluative coding [41]. The researchers examined the conceptual meanings of all responses and systematically coded the data. Codes were then grouped into broader categories, from which further subthemes were derived [40]. A comprehensive code list was prepared and organized into themes and subthemes. Any disagreements were discussed and resolved during team meetings. Regular meetings were held to further discuss and develop emerging codes, subthemes, and overarching themes. During these collaborative sessions, similar codes were merged, while distinct ones were extracted to provide a nuanced and detailed analysis of the data.

Methodological rigor: trustworthiness

The credibility of the study was ensured through several measures. (1) Purposive sampling was employed, selecting nurses from different departments, hospitals, and specialties to provide a diverse range of perspectives. (2) Efforts were made to ensure that the selected nurses did not know each other, reducing the potential for bias in their responses. (3) All interviews were audio-recorded, carefully listened to, and subsequently discussed by four researchers to ensure accuracy and depth of understanding. (4) Clear, concise, and pointed questions were asked during the interviews to elicit detailed and relevant information. (5) The analysis process involved four researchers who collaboratively discussed and validated the results. (6) All interviews were conducted by researchers with extensive experience and certification in research methodologies, further contributing to the study’s credibility. Finally, the study adhered to the SRQR checklist to ensure comprehensive and transparent reporting of the research [35]. These measures collectively contributed to the methodological rigor and credibility of the research findings.

Results

A total of 25 nurses participated in the study, providing a comprehensive overview of their backgrounds and experiences, as detailed in Table 1. The mean age of the participants was 28.32 years (SD = 3.42). The majority of the participants were women (92%) and single (64%). Educationally, 72% of the nurses held bachelor’s degrees in nursing, reflecting a wellqualified sample group. The participants had an average of 6.67 years of professional nursing experience (SD = 3.93). Their experience working in hospitals averaged 4.62 years (SD = 2.57), and their experience within their current units averaged 3.49 years (SD = 2.34). This varied range of experience provided a rich context for exploring the integration of AI in nursing practice. Employment status showed that 64% of the nurses were in permanent positions. In terms of work environment, 36% were employed in intensive care units, highlighting a critical care perspective, while 84% worked in public hospitals, offering insights into the public healthcare system.

Table 1.

Sociodemographic and professional characteristics of participants

Variables		Number of participants	Percentage, %
Gender	Female	23	92
	Male	2	8
Age in Years (Mean ± SD)		28.32 ± 3.42
Marital Status	Married	9	36
	Single	16	64
Education Level	Bachelor’s Degree Others	18 7	72 28
Professional Experience (years, Mean ± SD)		6.67 ± 3.93
Hospital Experience (years, Mean ± SD)		4.62 ± 2.57
Unit Experience (years, Mean ± SD		3.49 ± 2.34
Employment Status	Permanent	16	64
	Charge Nurse	9	36
Work Environment	Private Hospitals	4	16
	Public Hospitals	21	84

The qualitative analysis of the interviews identified four main themes and several related sub-themes (Table 2). These themes reflect nurses’ conceptualizations of AI, the perceived benefits and drawbacks of its integration into nursing practice, and their ethical and emotional responses to its use in clinical settings. The detailed thematic analysis, based on participant narratives and deeper sub-coding, provides valuable insights into the integration of AI into nursing, reflecting both the opportunities and concerns voiced by the participants.

Table 2.

Overview of themes and subthemes identified in nurses’ perspectives on the integration of artificial intelligence in nursing practice

Themes	Sub-themes	Illustrative codes / Concepts
Basic Understanding of Artificial Intelligence (AI)	Perceptions of AI Terminology	Robots, coded algorithms, machines with human-like reasoning
	Broader Conceptualization of AI	AI encompasses digital systems, software, and intelligent devices beyond robotics
Opportunities of AI in Nursing Practice	Workload Reduction and Resource Allocation	Automation of documentation, task delegation, and streamlined workflow
	Enabling Relational Nursing	More time for emotional support and interpersonal interaction
	Augmenting Clinical Procedures and Documentation	Use of AI in vital sign monitoring, treatment administration, and dosage regulation
	Support in Labor-Intensive Tasks	Robotic assistance in mobility, hygiene care, and positioning
Threats of AI in Nursing Practice	Job Displacement and Professional Redundancy	Reduction in nurse demand, fear of losing roles to automation
	Emotional Disconnection in Care	Absence of empathy, loss of therapeutic presence, concern over mechanized interactions
	Overdependence on Technology	De-skilling of nurses, underutilization of clinical judgment
	Dehumanization of Healthcare	Perceived transformation of caregivers into machine operators, rigid workflow patterns
Ethical and Psychological Concerns in AI-Based Nursing Practice	Ambiguity in Responsibility and Accountability	Unclear attribution of errors in AI-related incidents, legal uncertainty
	Risks to Patient Safety	Malfunctions, delayed responses, or errors in AI-led procedures
	Data Privacy and Cybersecurity Risks	Concerns over hacking, unauthorized access, and misuse of patient data
	Unsuitability for Psychiatric Settings	Risk of exacerbating psychiatric symptoms, lack of adaptive communication
	Psychological Discomfort Among Staff	Feeling of constant surveillance, reduced autonomy in care delivery.
	Potential for Misuse and Ethical Misconduct	Use of AI for malicious purposes, systemic bias, lack of transparency

Theme 1: the basic concepts of AI

All participant nurses had prior knowledge of AI. This theme includes two sub-themes: Perceptions of AI Terminology and Broader Conceptualization of AI. Most nurses associate the concept of AI primarily with robots. They focused on the algorithms that underpin AI, rather than on ML or machines acquiring human-like behaviors. For instance, one nurse stated,

“Not machines, but robots and algorithms loaded with technical hardware and codes.” (Maternity nurse, 30 years old).

This viewpoint underlines a technical understanding of AI, emphasizing the role of sophisticated programming and hardware in AI development.

Another nurse offered a broader perspective, distinguishing AI from purely robotic technology. She viewed AI as encompassing a range of computer systems that streamline nursing tasks, thus enhancing efficiency and care delivery. She remarked,

“Artificial intelligence encompasses all computers, technological devices, etc., around us. I think like that. While many people think of AI solely as robots, it is important to recognize that AI includes much more than just robotics.” (Physical therapy nurse, 26 years old).

This statement highlights the integration of AI in various technological devices used in daily nursing practices, from electronic health records to diagnostic tools. These perspectives illustrate the varied understanding of AI among nurses, highlighting a predominant association with robotics and coded algorithms while also acknowledging the broader scope of AI technologies in facilitating nursing work. This multifaceted understanding is crucial as it reflects the potential of AI to transform nursing practice through both technological advancements and improved patient care processes.

Theme 2: perceived opportunities of AI integration in nursing practice

Most nurses indicated that AI would significantly alter the roles of nurses. This theme reflects the perceived facilitators and positive implications of AI integration in clinical settings. Nurses described various dimensions of professional enhancement, with narratives organized into four sub-themes: (2.1) Workload Reduction and Resource Allocation, (2.2) Enabling Relational Nursing, (2.3) Augmenting Clinical Procedures and Documentation, and (2.4) Support in Labor-Intensive Tasks.

Workload reduction and resource allocation

Participants frequently cited AI’s potential to reduce the intensity of their workload, especially in high-pressure units such as intensive care units or emergency services. By automating routine tasks (e.g., vital sign monitoring, documentation), AI was seen as a tool for optimizing time and improving task distribution. This perception reflects the logic of operational efficiency that underpins many healthcare innovations.

“With AI, we could manage more patients. If I now take care of three patients, I might be able to handle four or five.” (Neurology ICU nurse, 27 years old).

Qualitatively, this suggests a perceived redefinition of the nursing role, from task executor to care coordinator, when supported by AI tools. However, it also highlights concerns about labor devaluation and role displacement.

Enabling relational nursing

A dominant narrative among participants was the hope that AI might “return time” to nurses, allowing them to engage more deeply in patient-centered, communicative, and empathetic care. Nurses described emotional labor and interpersonal connection as core components of professional satisfaction, elements often compromised by administrative burdens.

“If AI handles some of my workload, I can better communicate with patients. That improves both my job and their experience.” (Neurology ICU nurse, 30 years old).

This sub-theme emphasizes the importance of designing AI integration strategies that preserve and enhance the relational aspects of nursing, ensuring that technology does not depersonalize the care process but instead augments human interaction.

Augmenting clinical procedures and documentation

Several participants described AI as a facilitator of procedural accuracy and documentation efficiency. From automated dose calculations to streamlined digital charting, AI was perceived as a partner in clinical precision. The narratives indicated a belief that AI could serve as both a cognitive aid and risk mitigation tool.

“In pediatric clinics, adjusting drug dosages is difficult. With AI, you can input the drug and it prepares the dose and settings.” (Physical therapy nurse, 26 years old).

This viewpoint aligns with the ongoing digital transformation of clinical work and signals an emergent trust in algorithmic tools as support structures rather than threats to autonomy. A recent qualitative study on neonatal nurses working in Saudi Arabian NICUs confirmed that generative AI systems enhanced clinical decision-making and workflow efficiency, although successful implementation depended on robust training, infrastructure, and organizational readiness [42]. These findings resonate with our participants’ views on AI’s potential to augment procedural accuracy and ease physical burdens.

Support in Labor-Intensive tasks

AI was also described as a possible relief for physically demanding and repetitive tasks, such as lifting, hygiene, and transport. These insights particularly surfaced in accounts from nurses working in geriatric, long-term care, or intensive settings.

“AI can help us with things like bed baths, which are tough physically. But I’m not sure what problems might also come up.” (ICU nurse, Ph.D. student, 27 years old).

These accounts highlight not only the ergonomic potential of AI integration but also a latent ambivalence, nurses welcomed assistance but expressed concern about over-reliance or unforeseen consequences. This reflects a common dialectic in qualitative health research: appreciation for innovation tempered by ethical caution.

Theme 3: threats of AI in nursing practice

This theme explores the perceived threats that AI integration may pose to core aspects of nursing identity, human-centered care, and professional viability. While not always overtly hostile toward AI, participants critically reflected on how the technology might disrupt fundamental elements of their practice. Four sub-themes were derived: Job Displacement and Professional Redundancy, Emotional Disconnection in Care, Overdependence on Technology, and Dehumanization of Healthcare.

Job displacement and professional redundancy

Nurses were concerned that AI could reduce the need for human labor in healthcare, thereby threatening their employment and diminishing the role of the nurse.

“AI might lead to fewer nurses being needed. If machines do our tasks, our profession will shrink.” (Nursing Home Nurse, 29 years old).

This sub-theme reflects a broader anxiety about automation and its implications for workforce stability in the healthcare sector.

Emotional disconnection in care

Participants emphasized that AI cannot empathize, listen, or emotionally connect with patients, qualities that are central to practical and compassionate nursing.

“A robot can’t understand pain or emotions. We offer more than treatment; we provide comfort.” (Inpatient Nurse, 26 years old).

This critique challenges the assumption that technical competence alone is sufficient in caregiving, reinforcing the unique human qualities inherent in nursing.

Overdependence on technology

Some nurses feared that reliance on AI might result in a decline in clinical reasoning and manual skills, especially among less experienced staff.

“If we rely too much on AI, we may forget how to make our own clinical decisions.” (Physical Therapy Nurse, 26 years old).

This observation aligns with educational literature that cautions against the “deskilling” effects that can occur when clinicians excessively defer to automated systems.

Dehumanization of healthcare

A recurrent theme was the concern that AI might reduce nurses to machine operators, bound by rigid protocols and detached from the patient experience.

“We risk becoming button-pushers, following what the machine tells us, instead of caring professionals.” (ICU Nurse, Ph.D. student, 27 years old).

This sub-theme suggests a potential erosion of the caring ethos and relational aspects of nursing, replaced by technocratic efficiency and depersonalized workflows.

Theme 4: ethical and psychological concerns in AI-based nursing practice

This theme captures nurses’ multifaceted concerns regarding the ethical implications and psychological impact of AI use in healthcare settings. The integration of AI technologies presents not only logistical and clinical challenges but also deep-seated anxieties related to professional responsibility, legal ambiguity, patient safety, and emotional well-being in the workplace. Six sub-themes were identified: Ambiguity in Responsibility and Accountability, Risks to Patient Safety, Data Privacy and Cybersecurity Risks, Unsuitability for Psychiatric Settings, Psychological Discomfort Among Staff, and Potential for Misuse and Ethical Misconduct.

Ambiguity in responsibility and accountability

Nurses expressed uncertainty about who should be held accountable in cases where AI systems make errors or cause harm. While traditionally nurses assume responsibility for direct care actions, the introduction of autonomous systems disrupts this clear attribution of liability.

“If a robot measures a patient’s blood pressure and causes harm, I believe I shouldn’t be the one held responsible. The fault would lie with the system or the manufacturer.” (Neurology ICU Nurse, 26 years old).

This concern aligns with broader ethical discussions in healthcare AI regarding responsibility gaps, highlighting a pressing need for legal and institutional frameworks that define accountability in AI-assisted interventions.

Risks to patient safety

Participants voiced apprehension about the reliability and predictability of AI in critical care tasks. While AI is often associated with precision, any malfunction or delay in action can pose serious risks to patient well-being.

“AI may be helpful, but what if it malfunctions during an emergency? A delay could cost a life.” (ICU Nurse, Ph.D. student, 27 years old).

This sub-theme emphasizes the demand for rigorous validation and monitoring of AI tools in clinical environments to ensure safe and practical integration.

Data privacy and cybersecurity risks

Several nurses articulated anxiety about data breaches and the vulnerability of patient information in AI-operated systems. They questioned the security of digital infrastructures and worried about the ethical use of sensitive data.

“Even our personal computers can be hacked. I don’t think patient data stored in robots will be any safer.” (Neurology ICU Nurse, 30 years old).

These statements emphasize the ethical necessity of implementing robust cybersecurity protocols and transparent data governance in AI applications. Similar ethical dilemmas regarding AI recommendations conflicting with patient autonomy and cultural values were highlighted in a recent study conducted in Saudi Arabia, emphasizing the importance of transparent data-sharing practices and continuous ethics education for nurses in AI-integrated environments [43].

Unsuitability for psychiatric settings

Participants expressed concern that AI might be ill-suited for emotionally complex contexts, such as psychiatric care. They highlighted the need for adaptive human communication in these settings, something AI cannot yet replicate.

“In psychiatric clinics, patients often rely on empathy and eye contact. Robots can’t understand emotional cues or respond appropriately.” (Inpatient Nurse, 26 years old).

This sub-theme reflects a perceived mismatch between the emotionally nuanced needs of psychiatric patients and the mechanical, non-intuitive nature of current AI systems.

Psychological discomfort among staff

Some nurses mentioned that working alongside AI created a sense of surveillance and reduced professional autonomy. This constant monitoring and systematization of tasks contributed to stress and job dissatisfaction.

“It feels like everything we do is being watched or recorded. I miss having freedom in how I handle care.” (Nurse, 27 years old).

This finding resonates with sociotechnical critiques of digital health, which argue that excessive automation can erode clinicians’ sense of agency and professional judgment.

Potential for misuse and ethical misconduct

Finally, concerns were raised about the misuse of AI technologies, either intentionally (e.g., surveillance or biased decisions) or through systemic flaws in their design.

“There’s always a risk AI could be misused, like making decisions based on incomplete or biased data.” (Operating Room Nurse, 31 years old).

From a bioethical perspective, Armitage (2025) applies the principlism framework to large language models and argues that their integration may become ethically obligatory if risks are sufficiently mitigated, highlighting beneficence, autonomy, and justice as key pillars in responsible deployment [44]. These dimensions intersect directly with nurses’ concerns about explainability, patient consent, and equitable care delivery. Thus, this sub-theme emphasizes the importance of incorporating ethical principles into AI system design and ensuring transparency and oversight in AI deployment.

Discussion

Our research findings indicate that nurses are aware of AI and can anticipate advancements in AI technologies, as illustrated in Fig. 1.

Fig. 1.

Word cloud illustrating key themes from the participants’ perspectives as nurses in this study

Nurses identified numerous impacts of AI on healthcare and nursing services, including the automation of documentation, delegation of routine tasks, and optimization of workflows. Importantly, the sociocultural context in Turkey may shape nurses’ perceptions of AI in distinct ways. Cultural norms surrounding hierarchy in healthcare settings, expectations of emotional labor, and sensitivities to surveillance may have influenced participants’ concerns about professional autonomy, dehumanization of care, and ethical accountability. These contextual factors should be considered when interpreting the perceived threats and ethical dilemmas highlighted by the nurses. They also highlighted that AI could allow more time for emotional support and interpersonal communication with patients. Additionally, the use of AI in vital sign monitoring, treatment administration, and medication dosage regulation was recognized as beneficial. Robotic assistance in physically demanding tasks, such as patient mobility, hygiene care, and positioning, was also noted as a significant advantage. During the digital transformation journey, empowering nurses is essential because they continuously interact with patients and can effectively observe patient needs [14]. Nurses can play an important role in guiding the evolution of AI in nursing practice, particularly through their firsthand knowledge of patient care needs and clinical workflows [13]. The nursepatient interaction is crucial in guiding the development of AI and robotic applications by those who create these systems.

AIdriven applications such as robotics have the potential to provide essential services, assist patients with poor psychomotor skills, and transport obese or immobile patients. They can also reduce medication errors, especially in specialized departments like chemotherapy or pediatrics, and monitor vital signs using smart humanoid robots [45]. Health professionals can utilize these systems primarily for diagnosing and screening diseases such as cardiac diseases, coronary artery disease, myocardial infarction, and rhythm imaging devices like electrocardiograms [46]. Robotic applications in surgery show promise for further developments in healthcare. Studies suggest that robots could benefit immobile patients and provide geriatric care. The need for robots in healthcare arises in response to timesaving measures, patient safety, cost reductions, and the increasing number of individuals requiring healthcare [14]. AIbased robotic technologies can significantly impact the future of nursing and nursing management and education.

Participant nurses in this study also expressed concerns about the potential threats of AI, including the development of emotionally responsive robots, a reduced demand for nurses, and fears of role displacement due to automation. They emphasized the lack of empathy in AI systems, the loss of therapeutic presence in patient care, and concerns about increasingly mechanized interactions. Additionally, participants noted the risk of de-skilling among nurses and the underutilization of their clinical judgment. Some also perceived a troubling shift in the caregiver role, from compassionate professionals to machine operators constrained by rigid workflows. Furthermore, the emotional dimensions of nursing practice merit deeper analysis in the context of AI integration. As highlighted in our findings, several nurses expressed concerns about emotional disconnection in care, fearing that AI-driven systems may depersonalize interactions and diminish the therapeutic presence of nurses. A recent qualitative study conducted in 2025 with Jordanian nurses recognized the role of artificial intelligence in enhancing efficiency and reducing administrative burden. However, participants voiced ethical concerns related to patient privacy, the integrity of clinical decisions, and the potential decline in compassionate nurse–patient relationships [47]. These apprehensions mirror those expressed in our study, particularly regarding the loss of emotional connection in care delivery. This is consistent with the findings of Diogo et al. (2021), who emphasized that emotional competence constitutes a core element of nurses’ professional identity and the quality of care, especially in high-intensity clinical settings [19]. The most controversial issues related to AI-assisted applications include the risks of endowing machines with emotions, the possibility of machines operating outside human control, and the reduced need for human labor. Discussions about AI replacing humans in various jobs have been ongoing. A McKinsey report from June 14, 2023, highlighted the transformative potential of Generative AI (GenAI) in both business and societal contexts [48]. GenAI was identified as a crucial driver for the next wave of productivity, capable of unlocking novel capabilities and significantly enhancing existing functionalities across diverse areas, including imagery, video, audio, and coding. The year 2023 marked a landmark for generative AI, with applications such as OpenAI’s ChatGPT, Google’s Bard, Gemini, GitHub Copilot, and Stable Diffusion emerging prominently [9, 10]. These developments represented a significant milestone in AI’s evolution, demonstrating superior capabilities in processing extensive and diverse unstructured data. However, the McKinsey Global Institute (2017) report sparked debates about the positive and negative effects of AI globally, estimating that by 2030, 75 million to 375 million workers worldwide may need to change jobs due to AI technologies [49]. Researchers from the American Economic Association predict that while some tasks can be automated, few jobs can be entirely carried out by AI-based robots. The literature suggests that AI robots will only partially replace nursing roles in patient care, as skills such as communication, problem-solving, decision-making, and emotional intelligence are fundamental to nursing [14]. Hence, the use of robots for continuous care should be approached with skepticism. A study in a psychiatric hospital found that robot use changed patient behaviors, with some patients showing increased smiles and others displaying agitation [50]. Another study indicated that patients welcomed social robots, which were perceived as helpful in meeting their psychosocial needs [51]. Our findings echo this concern, particularly concerning AI technologies that may reduce nurses to mechanical roles or task executors. The risk of diminishing emotional labor, while often unrecognized in care metrics, could ultimately impact patient satisfaction and nurse retention. Therefore, future AI implementation strategies should not only prioritize efficiency and task delegation but also recognize and support the affective and relational components of nursing.

Nurses emphasized the ethical and social implications of AI, including the unclear attribution of responsibility in AI-related errors and associated legal uncertainties. Concerns were raised about potential malfunctions, delayed responses, and inaccuracies in AI-driven procedures. Participants also expressed anxiety over data privacy and cybersecurity, citing risks of hacking, unauthorized access, and misuse of patient information. Recent qualitative evidence published in 2024 from Turkey, health professionals emphasize that data privacy, ethical management of patient information, and the preservation of individualized care are key concerns surrounding AI-based decision support systems [52]. Similarly, in another 2024 qualitative study, in Saudi Arabia reported that 55% of nurses expressed ethical concerns, particularly related to patient privacy [53]. These findings underline the necessity of implementing governance frameworks and institutional strategies tailored to nurses, which ensure transparency, reinforce professional values, and foster trust in the integration of AI into healthcare practice. Additional concerns included the potential for AI to exacerbate psychiatric symptoms due to its lack of adaptive communication, feelings of constant surveillance among staff, and the erosion of professional autonomy. Finally, nurses highlighted the risk of AI being used for malicious purposes, systemic bias in algorithms, and a general lack of transparency in how AI systems operate. These concerns highlight the urgent need for clear ethical frameworks and regulatory guidelines that define accountability in AI-supported clinical decision-making [54, 55]. Without established protocols, both legal ambiguity and moral distress among nurses may increase. The perception of constant surveillance and reduced autonomy aligns with broader critiques of AI as a tool of managerial control, potentially shifting nursing from a caring profession to a data-monitored service role. These risks undermine the humanistic foundations of nursing practice. Moreover, the lack of transparency in AI algorithms raises ethical concerns regarding fairness, mainly when system decisions cannot be easily explained or justified. As previous studies have shown, such opacity may erode trust among healthcare providers and patients alike. According to the American Nurses Association, nurses are responsible for patient care, the nursing process, and interventions [56]. Technologies that assist in diagnosis, treatment, and care are meant to complement these roles, not replace them. The development of AI technologies in nursing emphasizes the importance of core nursing values and raises meaningful discussions on patient care ethics [57]. Furthermore, AI-based technologies prompt questions about patient privacy and individuality [18, 20]. Regarding transparency, accountability, and safety risks associated with AI algorithms, nurses who use these technologies may have greater confidence in their professional roles [14]. Nurses can maintain their clinical skills by adhering to ethical standards, making decisions, and monitoring AI behaviors [58]. They can also analyze challenges such as recognizing AI technologies, evaluating their accuracy and reliability, collaborating with experts for integration with sensor networks, ensuring patient safety, and managing human-machine interactions [38]. Considering the ethical dilemmas in literature, nurses should have knowledge and responsibility in controlling and managing AI technologies [57]. Every new technology presents both risks and opportunities. Nurses should stay informed about AI developments, participate in developing these applications, and propose beneficial effects for patient care. They should be encouraged to pursue relevant training at the undergraduate or graduate level [16].

Implications for policy and practice

This study revealed that nurses possess a fundamental understanding of AI concepts and recognize both the opportunities and the potential threats, including ethical and psychological concerns, that AI presents to their profession. Nurses expressed optimism about AI’s potential to reduce their workload, enhance patient imaging, diagnosis, transportation, and treatment processes. However, they also raised significant concerns regarding ethical and social dilemmas, particularly in terms of accountability, data security, and the irreplaceable value of human empathy and judgment.

For the policy lens, AI policies emphasized the complex nature and challenges of managing this technology. The most notable development was the EU AI Act, representing the first comprehensive legal framework for AI, which is scheduled to be published in 2024 and fully implemented by 2026 [59]. These steps are critical in shaping the future direction of AI, highlighting the increasing importance of political and ethical regulations post2024. Therefore, it is essential to develop comprehensive guidelines addressing the ethical use of AI in nursing, ensuring robust data security, clear accountability, and respect for patient privacy. Policymakers should establish regulatory standards that mandate regular monitoring and evaluation of AI systems in healthcare to safeguard against malfunctions and misuse. Additionally, policies should support continuous professional development programs that keep nurses updated on the latest AI technologies and their applications in healthcare.

From a practical perspective, interdisciplinary collaboration among nurses, managers, AI developers, and other healthcare professionals should be actively promoted to ensure that AI tools are designed and implemented in alignment with the specific needs of nursing practice. It is essential to encourage the use of AI as a complement rather than a replacement for the human elements of nursing care, thereby preserving the integrity of the nurse-patient relationship. By addressing these considerations, the integration of AI into nursing can be optimized to enhance operational efficiency, improve the quality of patient care, and uphold ethical standards, ultimately benefiting both nurses and patients within the evolving healthcare landscape.

Limitations

While our study offers valuable insights into the perceptions and implications of AI in nursing practice, it is essential to acknowledge the limitations inherent in qualitative research methodologies. Firstly, the findings are context-specific, which may limit the generalizability of the results. However, we have provided a detailed description of the study settings and participant demographics to support the potential transferability of our findings. Secondly, as our study employs a qualitative descriptive approach, any inferences drawn should be interpreted with caution. The subjective nature of qualitative data means that the findings reflect the personal experiences and opinions of the participants, which may not be universally applicable. Thirdly, the study captures the experiences of nurses within a specific local context in Istanbul, which may influence the interpretation of AI in nursing practice. The applicability of these findings may vary across different regions and healthcare systems. Another limitation is the exclusive focus on relatively young nurses. This narrow demographic scope may limit the generalizability of the findings, as it does not capture the perspectives of more experienced nurses who may approach AI technologies with greater critical insight grounded in long-term clinical practice. Veteran nurses often possess deeper familiarity with system-level operations and may be more attuned to both the limitations and unintended consequences of technological integration. Their absence may skew the results toward more optimistic or accepting views of AI. Additionally, the study relies on selfreported data, which may be subject to bias. Moreover, participants’ willingness to fully disclose their views may have been influenced by social desirability bias or concerns about professional repercussions.

Recommendations for future research

Future research can build upon our study by exploring the nuanced impacts of AI on various aspects of nursing practice. The field would benefit from inferential studies that examine the causal and predictive relationships between AI integration, nursing workflows, and patient outcomes. Such studies could provide a more detailed understanding of how AI technologies influence specific nursing tasks and overall care quality.

Additionally, there is a need for longitudinal research to assess the long-term effects of AI on nursing practice. While our study highlights immediate concerns and benefits, understanding the enduring impacts of AI implementation will require extended observation and analysis. This could include evaluating how AI affects job satisfaction, patient care quality, and the evolving roles of nurses over time. Further research should also consider the ethical and social dimensions of AI in nursing, with a focus on issues such as data security, accountability, and the emotional aspects of patient care. Investigating these areas can help develop comprehensive guidelines for the ethical use of AI in healthcare settings.

Moreover, comparative studies across different healthcare systems and cultural contexts would provide valuable insights into the variability of AI’s impact on nursing practice. Such research could identify the best practices and potential pitfalls in the global implementation of AI in nursing. Ultimately, interdisciplinary research that involves collaboration among nurses, nurse managers, AI developers, ethicists, and policymakers is crucial. This approach can ensure that AI technologies are developed and implemented in a manner that aligns with the needs and values of the nursing profession and the broader healthcare community. Therefore, future research should aim to include a more diverse nursing population in terms of years of experience, age, and professional roles (e.g., senior staff nurses, nurse managers, and clinical educators). Including experienced nurses could provide a broader and more nuanced understanding of AI integration, particularly regarding long-term implications, critical reflections, and organizational readiness. Their insights may help identify barriers and enablers that are not evident among younger or less experienced staff.

Conclusion

This study provides valuable insights into nurses’ perspectives on integrating artificial intelligence (AI) into nursing practice. Participants acknowledged that AI can be utilized in various aspects of nursing, including automating documentation, supporting clinical decision-making, aiding in patient monitoring, and enhancing workflow efficiency. However, they emphasized that AI should complement, not replace, the human elements of care, particularly the emotional and interpersonal dimensions that define the nursing profession. Nurses expressed both optimism and caution regarding the future role of AI in nursing. While recognizing its potential to enhance care quality and reduce workload, they also raised critical concerns about ethical and social implications. Many participants highlighted the ambiguity around legal accountability and transparency in AI systems, reinforcing the need for strong ethical guidelines. Importantly, nurses envisioned AI as a powerful tool if integrated thoughtfully and responsibly. They predicted that AI will become increasingly important in healthcare and stressed that its successful implementation depends on interdisciplinary collaboration, adequate training, and nurse involvement in decision-making processes. To effectively integrate AI-based technologies into nursing care and practice, nurses must possess a foundational understanding of these systems. For AI to benefit both patients and providers, it must align with the core values of nursing, compassion, empathy, and human connection, while ensuring ethical integrity and promoting professional development. Nurses should actively engage in research and development processes, embracing AI technologies rather than resisting them, to evaluate their potential benefits and ensure their use aligns with ethical principles, while also being aware of associated risks.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Study design: GB, AKHS Data collection: HÇ, ST Data analysis: GB, HÇ, ST, PG Study supervision: AKHS Manuscript writing: GB, HÇ, ST, AKHS, PG Critical revisions for important intellectual content: GB, AKHS, PG.

Funding

The authors have not received any funding from public or private institutions to support the conduct of this study.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the Istanbul University-Cerrahpasa, Social and Human Sciences Ethical Committee (date: 4 October 2021, code: 2021/204). The study was conducted in accordance with the principles of the Declaration of Helsinki. Before the interviews, participants were provided with an information sheet detailing the study objectives and procedures. Written and verbal informed consent was obtained prior to recording, and participants confirmed their willingness to join the study by selecting the “I agree to participate” option at the beginning of the interview. They were informed of their right to refuse any question or to withdraw from the study at any time, in line with the principles of informed consent and respect for autonomy. Confidentiality was ensured, and all personal information was kept strictly anonymous. Both the participants and the researcher retained copies of the signed consent forms to ensure transparency and compliance with ethical standards.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.