Facilitators and barriers to AI adoption in nursing practice: a qualitative study of registered nurses' perspectives

Abstract

Background

Integrating Artificial Intelligence (AI) in nursing practice is revolutionising healthcare by enhancing clinical decision-making and patient care. However, the adoption of AI by registered nurses, especially in varied healthcare settings such as Saudi Arabia, remains underexplored. Understanding the facilitators and barriers from the perspective of frontline nurses is crucial for successful AI implementation.

Aim

This study aimed to explore registered nurses' perspectives on the facilitators and barriers to AI adoption in nursing practice in Saudi Arabia and to propose an extended Technology Acceptance Model for AI in Nursing (TAM-AIN).

Methods

A qualitative study utilising focus group discussions was conducted with 48 registered nurses from four major healthcare facilities in Al-Kharj, Saudi Arabia. Thematic analysis, guided by the Technology Acceptance Model framework, was employed to analyse the data.

Results

Key facilitators of AI adoption included perceived benefits to patient care (85%), strong organisational support (70%), and comprehensive training programs (75%). Primary barriers involved technical challenges (60%), ethical concerns regarding patient privacy (55%), and fears of job displacement (45%). These findings led to the development of TAM-AIN, an extended model that incorporates additional constructs such as ethical alignment, organisational readiness, and perceived threats to professional autonomy.

Conclusions

AI adoption in nursing practice requires a holistic approach that addresses technical, educational, ethical, and organisational challenges. The proposed TAM-AIN offers a comprehensive framework for optimising AI integration into nursing practice, emphasising the importance of nurse-centred implementation strategies. This model provides healthcare institutions and policymakers with a robust tool to facilitate successful AI adoption and enhance patient outcomes.

Introduction

The rise of Artificial Intelligence (AI) in healthcare represents a transformative milestone in the evolution of medical practice, with the potential to dramatically reshape patient care, clinical decision-making, and healthcare management [1, 2]. AI, broadly defined as the simulation of human intelligence processes by machines, encompasses a variety of technologies such as machine learning, natural language processing, robotics, and computer vision [3–5]. In healthcare, AI applications range from diagnostic tools and personalised treatment strategies to automating administrative tasks and utilising predictive analytics for managing population health [6, 7]. However, in the context of this study, AI specifically refers to the deployment of these technologies to assist in clinical decision-making, patient monitoring, and workflow optimisation in nursing practice [8–10].

The potential of AI in nursing practice is significant and multifaceted [11]. AI-powered predictive analytics can assist nurses in identifying patients at high risk of deterioration, facilitating early interventions, and potentially reducing adverse events [12–15]. AI-driven patient monitoring systems continuously analyse physiological data, alerting nurses to subtle changes that might otherwise be overlooked [16, 17]. For example, AI algorithms integrated into electronic health records (EHRs) can automatically alert nurses to potential sepsis in patients, allowing for timely intervention [18, 19]. Another practical application is the use of AI in managing chronic conditions, where AI tools can analyse patient data over time to provide nurses with actionable insights for adjusting treatment plans, thus improving patient outcomes [20–22]. Additionally, AI-enhanced clinical decision support systems can offer evidence-based recommendations precisely when needed, supporting nurses in making accurate and timely decisions [23]. These advancements in AI hold the promise of significantly improving the quality of patient care and transforming nursing workflows [24].

Despite AI's promising potential, its adoption in nursing practice remains inconsistent and fraught with challenges [25–27]. The existing literature on AI in healthcare has predominantly focused on the technical capabilities and clinical outcomes associated with AI deployment [28, 29]. For instance, several studies highlight the impact of AI on clinical decision-making and patient outcomes, yet they often overlook the nuanced perspectives of nurses who are critical to the successful integration of AI technologies [30–33]. Additionally, much of the research has focused on physicians and healthcare administrators, leaving a significant gap in our understanding of how nurses, as frontline healthcare providers, perceive and interact with AI [34, 35].

This gap is particularly evident in the lack of qualitative research that delves into the experiences, concerns, and attitudes of nurses regarding AI adoption [36–38]. While quantitative studies have provided some insights into general attitudes toward AI, they often fail to capture the complex interplay of factors that influence these attitudes [39, 40]. For example, there is limited exploration of the ethical concerns that nurses might have about AI, such as its impact on patient privacy or the potential to erode the nurse-patient relationship [41]. Furthermore, there is a notable absence of studies examining how the cultural and healthcare context in specific regions, such as Saudi Arabia, influences AI adoption in nursing [42, 43].

Saudi Arabia's healthcare system is currently undergoing rapid modernisation, with significant investments in digital health initiatives [44–48]. However, the adoption of AI technologies is influenced by several factors unique to the region, including variations in technological infrastructure, cultural attitudes towards technology, and the availability of resources [49–51]. For example, large urban hospitals may have the capacity to implement advanced AI systems, while smaller rural healthcare facilities may struggle due to financial constraints and limited access to cutting-edge technology [52, 53]. Additionally, cultural factors, such as varying levels of comfort with new technologies among healthcare professionals, play a crucial role in how AI is perceived and integrated into nursing practice [54–57]. These contextual factors are critical to understanding the broader landscape of AI adoption in the region and tailoring strategies that are culturally and technologically appropriate [58–60].

To bridge the significant gaps identified in the literature regarding AI adoption in nursing, this study focuses on exploring the perspectives of registered nurses in Saudi Arabia. In this context, such insights are particularly scarce. Nurses, being the primary caregivers and largest group of healthcare professionals, play a pivotal role in the successful implementation of AI technologies in clinical settings. Understanding their views is essential for tailoring strategies that facilitate effective AI integration while addressing potential concerns. Therefore, the central research question guiding this study is:

• Q1. What are the perceived facilitators and barriers to the adoption of artificial intelligence in nursing practice among registered nurses in Saudi Arabia?

To comprehensively explore this question, we have set the following specific objectives:

1. To identify and analyse the key facilitators that promote the successful integration of AI technologies in nursing practice, as perceived by registered nurses. This involves uncovering factors that nurses believe contribute positively to AI adoption, such as perceived benefits to patient care, ease of use, and organisational support.

2. To elucidate and examine the primary barriers that impede the adoption and implementation of AI in nursing workflows from the perspective of nursing professionals. This seeks to understand the challenges nurses face in adopting AI, including technical difficulties, ethical concerns, and perceived threats to professional autonomy.

3. To gain a comprehensive understanding of nurses' concerns, expectations, and attitudes towards using AI in their professional practice. This objective explores the emotional and cognitive aspects of nurses' engagement with AI, encompassing their hopes, fears, and ethical considerations.

4. To develop evidence-based recommendations for enhancing AI adoption in nursing, informed by the insights and experiences of frontline practitioners. This aims to translate the study's findings into practical strategies for healthcare organisations, policymakers, and technology developers.

This study is significant because it centers on understanding registered nurses' perspectives on AI adoption within their practice. As the largest group of healthcare professionals and the primary providers of direct patient care, nurses are instrumental in the successful implementation of any new technology in healthcare settings [61]. However, existing studies often fail to capture the practical implications of AI for nursing practice [62]. For instance, while quantitative surveys have shed light on nurses' general attitudes toward AI, they often lack the depth needed to grasp the complex interplay of factors that shape these attitudes [63]. Additionally, research often overlooks why other stakeholder perspectives, such as those of patients or administrators, are not included despite their relevance to the broader conversation [64]. The focus on nurses in this study is deliberate, as their direct involvement in patient care makes their perspective particularly crucial. However, this does not negate the importance of exploring other stakeholder views in future research [65, 66]. AI technologies hold considerable potential for transforming nursing practice by enhancing patient care, improving efficiency, and supporting clinical decision-making [67, 68]. While promising, it is essential to recognise that the successful integration of AI into nursing practice requires careful consideration of its limitations, including technical challenges, ethical concerns, and the need for ongoing education and training [69–71]. By understanding both the opportunities and challenges AI presents, healthcare organisations can make more informed decisions regarding its implementation in nursing settings.

In conclusion, this research is paramount in decoding the intricate landscape of AI adoption in nursing practice, uniquely centered on the perspectives of registered nurses in Saudi Arabia. By illuminating both the catalysts and impediments to AI integration, it fills a critical gap in the current literature and provides invaluable insights for a diverse range of stakeholders, including policymakers, educators, and healthcare administrators. The findings will serve as a cornerstone for developing strategies that harmonise AI implementation with fundamental nursing values, thereby enhancing patient care while safeguarding the profession's indispensable human elements. In an era of unprecedented technological advancement in healthcare, this study's emphasis on the human dimensions of AI adoption is not only timely but crucial. It offers a roadmap for leveraging AI to augment and empower nurses, ensuring that technological progress amplifies rather than diminishes the essence of nursing practice.

Theoretical framework

This study will employ the Technology Acceptance Model (TAM) as a theoretical framework to guide the exploration of nurses' perspectives on AI adoption. Originally proposed by Davis (1989) [72] and later extended by Venkatesh and Davis (2000) [73, 74], the TAM provides a robust structure for analysing the factors that influence technology acceptance and use [75, 76]. The model posits that perceived usefulness and perceived ease of use are primary determinants of technology adoption, mediated by attitudes towards use and behavioural intentions [77, 78]. Figure 1 illustrates this theoretical framework, specifically applied to the context of AI adoption in nursing practice. By applying the TAM to the context of AI adoption in nursing practice, this study aims to provide a theoretically grounded analysis of the facilitators and barriers influencing nurses' acceptance of AI technologies. The TAM's construct will be used to structure the inquiry, helping to categorise and analyse the various factors that emerge from nurses' perspectives. However, recognising the unique context of nursing and the complex nature of AI technologies, this study will also remain open to identifying factors that may extend beyond the traditional TAM framework, potentially contributing to an expanded model of technology acceptance specific to AI in nursing (Fig. 2).

Fig. 1.

Theoretical Framework for AI Adoption in Nursing Practice Based on the Technology Acceptance Model (TAM)

Fig. 2.

Sample Procedure for AI Adoption Study in Nursing Practice

Materials and methods

Research design

This study employed a descriptive qualitative research design to explore registered nurses' perspectives on the facilitators and barriers to AI adoption in nursing practice [79–82]. The Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist was followed to ensure methodological rigour, and a completed checklist was included as supplementary material [83]. Focus groups were used as the primary data collection method, allowing for the exploration of shared experiences, group dynamics, and collective attitudes among registered nurses [84–87]. Each group comprised 6–8 participants, selected to ensure diversity in experience, specialities, and practice settings. Sessions were moderated by a trained facilitator using a semi-structured guide to encourage comprehensive discussions while exploring emergent themes. Thematic analysis was employed to analyse audio-recorded and transcribed data, integrating insights from focus group discussions, observers' field notes, and demographic information. Rigor was enhanced through strategies such as member checking, reflexive journaling, and peer debriefing. Data triangulation further validated findings by cross-referencing multiple data sources to ensure consistency and provide a comprehensive understanding of AI adoption in nursing.

Settings

This study was conducted in Al-Kharj, Saudi Arabia, across four diverse healthcare facilities selected to represent various nursing practice environments and explore facilitators and barriers to AI adoption comprehensively. The first site, a large general hospital with 246 beds and approximately 300 nurses, provided a broad range of services, including internal medicine, surgery, paediatrics, and rehabilitation. Its diverse specialities offered insights into AI adoption across acute care and rehabilitation contexts. The second facility, a specialised hospital serving Ministry of Defense personnel and their families, had 100 beds and around 150 nurses. It focused on emergency care, surgery, and general medical services, allowing for an exploration of AI integration in specialised healthcare settings. The third site, a maternal and child health hospital with 150 beds and approximately 200 nurses, was dedicated to obstetrics, paediatrics, and neonatal care, highlighting the role of AI in improving outcomes in maternal and child health nursing. Finally, the fourth site, a university hospital with 200 beds and about 250 nurses, offered outpatient clinics in various specialities and provided unique perspectives on AI adoption in nursing education and research as a teaching institution. These settings, varying in size, specialisation, and patient populations, facilitated a comprehensive examination of the challenges and opportunities associated with AI adoption across different healthcare contexts, ultimately informing targeted recommendations for its integration into nursing practice.

Sample

Purposive sampling was employed to recruit registered nurses from four hospitals in Al-Kharj, Saudi Arabia, ensuring the selection of information-rich cases to provide in-depth insights into AI adoption in nursing practice [88]. Inclusion criteria included: (a) current employment in one of the selected hospitals, (b) a minimum of one year of clinical experience, (c) familiarity with AI technologies ranging from basic awareness to active use, and (d) willingness to participate in focus group discussions. The sample size was guided by data saturation, achieved when no new themes or insights emerged from additional data collection [89]. Based on similar qualitative studies, we planned for eight focus groups of six participants each (48 total). This number was confirmed as sufficient as saturation was reached after analysing themes from all eight sessions [89]. Maximum variation sampling within the purposive framework ensured diversity across clinical experience (early, mid-career, senior), nursing specialities (e.g., medical-surgical, critical care), AI exposure levels (low, moderate, high), educational backgrounds (diploma, bachelor’s, postgraduate), age groups (20–50 + years), and gender. Recruitment was facilitated through collaboration with nursing administration departments via meetings, communications, and flyers. Interested nurses contacted the research team directly to maintain confidentiality and avoid coercion. Challenges in recruiting participants with low AI exposure were mitigated by emphasising the value of all perspectives. A recruitment log documented participant characteristics and reasons for non-participation, addressing potential sampling biases. Demographic and professional details, along with AI exposure levels, are presented in the results to support the credibility and transferability of the findings [90].

Data collection tools

Our primary research instrument was a semi-structured focus group interview guide designed to elicit rich [91], detailed perspectives from registered nurses on AI adoption in clinical practice. This guide was developed through an iterative process, informed by the Technology Acceptance Model (TAM) framework and a comprehensive review of contemporary healthcare AI literature. The interview guide explored four key domains: understanding of AI in nursing practice, perceived facilitators of AI adoption, potential barriers to implementation, and attitudes toward AI integration. Key questions included "What challenges or concerns do you face when considering the adoption of AI technologies?" and "How do you think AI might impact the core values and identity of the nursing profession?" Follow-up probes explored participants' responses, particularly regarding perceived usefulness and ease of use. The guide underwent rigorous validation through an expert panel review comprising nursing informatics specialists and qualitative researchers. Cognitive interviews with nurses assessed question clarity while pilot testing refined flow and timing. This structured development process resulted in a flexible instrument that progressed from general to specific questions, facilitating natural discussion while ensuring comprehensive topic coverage. Two supplementary instruments enhanced data collection: a demographic questionnaire gathering participant characteristics (age, gender, experience, education, speciality, and AI exposure) for cross-comparison analysis and a field notes template documenting group dynamics, non-verbal cues, and emerging themes during sessions. All instruments underwent back-translation between English and Arabic [92], carefully considering cultural nuances in question phrasing and group dynamics relevant to the Saudi Arabian context.

Ethical approval

This study was conducted in accordance with ethical standards for research involving human subjects. Approval was obtained from the Standing Committee of Bioethics Research (SCBR) at Prince Sattam bin Abdulaziz University, Deanship of Scientific Research (Approval No. SCBR-309/2024). The study protocol was reviewed to ensure compliance with the university's guidelines. Informed consent was obtained from all participants, ensuring they were fully aware of the study's purpose, procedures, potential risks, and their rights to withdraw at any time. Participant confidentiality was maintained through data anonymisation and secure storage, adhering to data protection regulations. These measures underscored our commitment to conducting rigorous, ethically sound research that respected participants' rights and welfare while contributing valuable insights to AI adoption in nursing practice.

Procedure

The study followed a systematic and rigorous approach to data collection and analysis. Data collection tools were developed based on the Technology Acceptance Model (TAM) framework, created in English, and back-translated into Arabic to ensure conceptual equivalence and cultural relevance. A pilot focus group was conducted to refine the interview guide and assess the feasibility of the data collection process. Recruitment, described in Section " Sample", ensured diverse perspectives through purposive sampling with maximum variation. Eight focus groups, each comprising six participants (totalling 48 nurses), were conducted in private rooms at the participating hospitals to ensure participant comfort. Sessions lasted 90–120 min and were moderated by a trained facilitator using a semi-structured interview guide. An observer documented group dynamics and emerging themes. All sessions were audio-recorded with participants' consent to ensure accurate data capture. Participants completed a demographic questionnaire before the discussions. Each session began with an introduction explaining the study’s purpose and a discussion of ground rules. Questions progressed from general to specific, covering key domains: understanding of AI, perceived facilitators, barriers, and attitudes toward AI adoption. Probing questions encouraged deeper exploration of responses aligned with TAM constructs. Post-session, facilitators and observers debriefed to identify initial impressions and emerging themes. Audio recordings were transcribed verbatim and reviewed for accuracy. Thematic analysis, guided by Braun and Clarke’s approach (2006) [93], combined deductive and inductive methods. Two researchers independently coded transcripts, with regular team meetings held to refine the coding framework. NVivo software was used to organise and manage qualitative data efficiently.

Data triangulation and quality assurance

Data triangulation was employed to enhance the credibility and robustness of the findings by integrating multiple data sources, including focus group transcripts, field notes, and demographic questionnaires [94]. These sources were cross-referenced during analysis to comprehensively understand facilitators and barriers to AI adoption in nursing practice. Investigator triangulation involved two researchers independently coding the data, followed by collaborative discussions to refine themes, minimise biases, and enhance the validity of the findings. Quality assurance measures included member checking, audit trails, and peer debriefing. Preliminary findings were shared with a subset of participants for validation (member checking), and an audit trail was maintained to ensure transparency and dependability. Peer debriefing sessions with uninvolved colleagues further challenged assumptions and interpretations, contributing to the rigour of the analysis. Ethical considerations, as outlined in Section " Ethical approval", were strictly adhered to throughout the study. These included informed consent, confidentiality, and compliance with approved protocols, ensuring the research was conducted ethically and aligned with best practices.

Data analysis

Thematic analysis was conducted to identify key themes emerging from the focus group discussions. This analysis followed the approach outlined by Braun and Clarke (2006) [93], combining both inductive and deductive coding to ensure a comprehensive data exploration. Two researchers independently coded the transcripts to ensure inter-rater reliability. The coding was then refined through collaborative discussions to develop the final thematic framework. While the thematic analysis focused on qualitative insights, we used simple counts and percentages to represent the prevalence of certain themes. This approach was adopted to enhance the clarity of the findings, allowing readers to easily grasp the relative frequency with which participants raised specific facilitators or barriers to AI adoption. The use of percentages here is not meant to imply quantitative analysis but rather to provide additional context for the thematic insights by indicating how commonly a particular theme appeared across the discussions. This combination of qualitative and quantitative elements provides a richer, more detailed understanding of the factors influencing AI adoption in nursing while ensuring that the analysis remains firmly grounded in qualitative methodology.

Results

Overview of the participants

A total of 48 registered nurses participated in the study, with a mean age of 36.7 years (SD ± 8.4) and an average of 11.3 years (SD ± 7.2) of nursing experience. The majority of participants were female (77.1%) and held a bachelor's degree (66.7%). Participants represented diverse specialities, including medical-surgical (37.5%), critical care (20.8%), and paediatrics (16.7%). AI exposure levels ranged from low to high, with the majority (45.8%) reporting moderate exposure. Table 1 provides an overview of the participants' demographic and professional characteristics.

Table 1.

Demographic and Professional Characteristics of Participants and Their Perspectives on AI Integration in Nursing Practice

Characteristic	N (%) or Mean ± SD
Age (years)	36.7 ± 8.4
Gender
Female	37 (77.1%)
Male	11 (22.9%)
Years of Experience	11.3 ± 7.2
Education Level
Diploma	10 (20.8%)
Bachelor’s Degree	32 (66.7%)
Postgraduate	6 (12.5%)
Nursing Specialty
Medical-Surgical	18 (37.5%)
Critical Care	10 (20.8%)
Pediatrics	8 (16.7%)
Obstetrics	6 (12.5%)
Other Specialties	6 (12.5%)
AI Exposure Level
Low	20 (41.7%)
Moderate	22 (45.8%)
High	6 (12.5%)

Key themes identified

Thematic analysis of focus group discussions identified four overarching themes and corresponding subthemes, which are summarised in Table 2. Figure 3 visually represents the interconnectedness of these themes.

Table 2.

Themes and Subthemes Generated in this Study

Themes	Subthemes
(1) Understanding of AI in Nursing Practice	1a. Basic understanding of AI
	1b. Misconceptions about AI capabilities
(2) Facilitators of AI Adoption	1c.Advanced understanding based on direct experience
	2a. Perceived benefits to patient care
	2b. Organizational support and leadership commitment
	2c. Ease of use and workflow integration
	2d. Educational and training programs
(3) Barriers to AI Adoption	3a. Technical challenges and resource limitations
	3b. Ethical and privacy concerns
	3c. Fear of job displacement and deskilling
	3d. Resistance to change and scepticism
(4) Attitudes Toward AI Integration	4a. Cautious optimism
	4b. Expectations for gradual implementation

Fig. 3.

Thematic Analysis of Key Domains Influencing AI Adoption in Nursing Practice (N = 48)

Theme 1: understanding of AI in nursing practice

Subtheme 1a: basic understanding of AI

The majority of participants (65%) displayed a foundational understanding of artificial intelligence (AI) and its applications in nursing practice. This understanding primarily revolved around AI’s ability to automate routine tasks, improve clinical decision-making, and analyse large sets of patient data efficiently. Participants commonly perceived AI as a supportive tool that complements their work by reducing workload and enhancing accuracy. One participant succinctly captured this perspective, stating,

AI to me means smart systems that can help us detect patient deterioration early or alert us about medication interactions (Participant 15, Female, 34 years).

Such statements illustrate that participants recognised AI’s practical utility in identifying clinical issues earlier, providing timely alerts, and assisting in decision-making processes that contribute to improved patient safety and care outcomes. This level of comprehension indicates a functional awareness of AI’s potential to optimise healthcare delivery while supporting nursing's core responsibilities. However, the emphasis was on its supplementary role, aligning with the nurses’ broader expectations of technology as a partner in clinical workflows rather than a standalone solution.

Subtheme 1b: misconceptions about AI capabilities

A notable proportion of participants (25%) exhibited misconceptions regarding the capabilities of artificial intelligence (AI) in nursing practice. These misconceptions often stemmed from a limited understanding of how AI integrates into clinical workflows. Participants expressed apprehensions about AI potentially replacing human nurses, which created a sense of unease about its role in healthcare. For example, one participant remarked,

I thought AI meant robots would completely replace nursing tasks, but now I understand it is more about supporting our decision-making (Participant 27, Male, 41 years).

This comment highlights a shift in perception once participants gained a clearer understanding of AI’s intended purpose. However, such initial misconceptions underscore the need for structured educational initiatives. Providing accurate information and clarifying AI's supportive role in nursing are essential to addressing these fears and fostering confidence in its integration into practice.

Subtheme 1c: advanced understanding based on experience

A smaller subset of participants (10%) demonstrated an advanced understanding of AI, typically linked to direct, hands-on experience with AI tools in their clinical environments. These participants articulated detailed insights into how AI technologies enhance specific aspects of care delivery, such as early detection of complications and timely decision-making. One participant reflected on their experience, stating,

Using AI in fetal monitoring systems has made detecting complications faster and more precise, enabling timely interventions (Participant 42, Female, 36 years).

This advanced understanding highlights the potential for experience-driven familiarity to build confidence in AI applications. These participants perceived AI not only as a tool for support but also as an integral component of optimising patient outcomes in high-stakes scenarios. Their experiences emphasise the value of practical exposure to AI in shaping positive attitudes and understanding among nursing professionals.

Theme 2: facilitators of AI adoption

Subtheme 2a: perceived benefits to patient care

A significant majority of participants (85%) emphasised AI's potential to enhance patient care by improving diagnostic accuracy, reducing human error, and enabling timely interventions. Participants noted that AI's advanced analytical capabilities allow for faster processing and interpretation of patient data, providing crucial insights that may not be immediately apparent through traditional methods. A critical care nurse highlighted this benefit, explaining,

AI systems can analyse patient data much faster than we can. This could help us identify deteriorating patients earlier, potentially saving lives (Participant 23, Female, 42 years).

Such observations reflect a shared belief that AI’s ability to detect subtle changes in clinical indicators can play a pivotal role in preventing adverse outcomes. Participants further noted that AI could complement their decision-making processes by providing real-time alerts and evidence-based recommendations, thereby supporting safer and more efficient patient management.

Subtheme 2b: organizational support and leadership commitment

Organisational support and leadership commitment were identified as critical factors in facilitating AI adoption by 70% of participants. Participants stressed that strong leadership involvement in promoting and investing in AI technologies fosters a culture of innovation and trust, making staff more receptive to integrating these tools into their practice. For instance, a nurse manager remarked,

When our hospital leadership actively promotes and invests in new technologies like AI, it creates a culture of innovation. This makes staff more open to embracing these tools (Participant 8, Female, 47 years).

Conversely, participants from facilities with limited leadership engagement reported challenges in obtaining the resources and encouragement needed to adopt AI effectively. These findings underscore the importance of visible, proactive leadership in shaping institutional readiness for AI integration.

Subtheme 2c: ease of use and workflow integration

Ease of use and seamless integration of AI tools into existing workflows were highlighted as essential facilitators by many participants. Participants consistently emphasised that user-friendly interfaces and compatibility with current systems, such as electronic health records (EHRs), are crucial for promoting AI adoption. When AI tools are intuitive and require minimal additional effort to operate, they are more likely to be embraced by nursing staff. A medical-surgical nurse explained,

If AI tools are intuitive and integrate well with our current electronic health records, we’re more likely to use them (Participant 31, Female, 33 years).

Participants also noted that poorly designed systems that add complexity or disrupt established workflows could discourage adoption, underscoring the need for tools that complement rather than complicate nursing practice.

Subtheme 2d: educational and training programs

The role of education and training in facilitating AI adoption was emphasised by 75% of participants. Nurses highlighted the importance of comprehensive training programs that equip them with the knowledge and skills to use AI effectively. Hands-on sessions, demonstrations of AI applications, and opportunities to address concerns were frequently mentioned as key elements of effective training initiatives. One nurse shared,

The training we received on our new AI-assisted fetal monitoring system was excellent. It made me feel confident in using the technology and understanding its benefits (Participant 42; Female, 36 years).

Participants also stressed the need for ongoing training to keep pace with advancements in AI technology, ensuring that staff remain competent and confident in using emerging tools. This highlights the critical role of continuous professional development in fostering positive attitudes toward AI adoption in nursing practice.

Theme 3: barriers to AI adoption

Subtheme 3a: technical challenges and resource limitations

A substantial proportion of participants (60%) identified technical challenges, particularly outdated IT infrastructure and insufficient resources, as significant barriers to AI adoption. Participants explained that many healthcare facilities lack the technological capacity to support advanced AI tools, often requiring comprehensive upgrades to existing systems. These upgrades were perceived as resource-intensive, both in terms of cost and implementation logistics. One participant expressed this concern succinctly:

Our current IT system is outdated. I worry that implementing AI would require a complete overhaul (Participant 5, Male, 39 years).

Nurses working in smaller or under-resourced facilities emphasised the disparity in technological readiness compared to larger, better-funded institutions. The lack of appropriate infrastructure hindered AI adoption and exacerbated frustrations among staff, who felt that technological advancements were inaccessible to them. Participants noted that overcoming these challenges would require targeted investment and strategic planning to ensure equitable access to AI technologies across different healthcare settings.

Subtheme 3b: ethical and privacy concerns

Ethical concerns related to patient data privacy and the secure handling of sensitive information were raised by 55% of participants. Nurses highlighted the potential risks associated with storing vast amounts of patient data in AI systems, particularly the possibility of breaches or misuse of information. Participants emphasised the importance of implementing robust security measures to protect patient confidentiality and ensure compliance with legal and ethical standards. One participant voiced this apprehension:

While I see the benefits of AI, I’m concerned about patient privacy. How can we ensure that sensitive data isn’t misused? (Participant 14, Female, 52 years).

Beyond privacy, participants also questioned the ethical implications of relying on AI for clinical decision-making, particularly in scenarios where algorithms might override human judgment. This highlights the need for clear guidelines and ethical frameworks to govern AI use in nursing, ensuring that patient welfare and professional integrity remain paramount.

Subtheme 3c: fear of job displacement and deskilling

Fear of job displacement and the potential erosion of nursing skills. These concerns were particularly pronounced among experienced nurses, who worried that as AI increasingly automates tasks, the value of critical thinking and hands-on skills might diminish. Participants expressed apprehension that over-reliance on AI could lead to a deskilled workforce, particularly among newer generations of nurses who may not have the same level of experience in traditional nursing practices. One experienced nurse explained,

I worry that as AI takes over more tasks, our skills might become less valued (Participant 33, Male, 50 years).

Despite these fears, some participants acknowledged that AI could serve as a tool to enhance rather than replace nursing roles, provided that its implementation is carefully managed. This dual perspective underscores the importance of balancing automation with opportunities for professional development and maintaining the human elements of nursing care.

Subtheme 3d: resistance to change and skepticism

Resistance to AI adoption was reported by 40% of participants, with scepticism about the technology’s reliability and its ability to replicate the nuanced judgment required in nursing. Nurses with limited exposure to AI were particularly hesitant, often questioning whether AI systems could truly understand the complexities of patient care. Participants emphasised that nursing involves technical skills, interpersonal connections, and empathy, which they felt AI could not replicate. A critical care nurse remarked,

I am not convinced that AI can truly understand the nuances of patient care. There is an art to nursing that machines cannot replicate (Participant 37, Male, 44 years).

This scepticism highlights the need for evidence-based demonstrations of AI’s reliability and effectiveness in clinical practice. Participants suggested that clear communication, practical examples, and success stories could help alleviate these concerns and build trust in AI technologies over time.

Theme 4: attitudes toward AI integration

Subtheme 4a: cautious optimism

70% of participants expressed a sense of cautious optimism regarding the integration of AI into nursing practice. While participants acknowledged AI’s potential to enhance efficiency, reduce workloads, and improve decision-making, they consistently emphasised the importance of maintaining the human element of nursing care. Nurses viewed AI as a valuable tool that could complement their expertise rather than replace their critical thinking and interpersonal skills. This balance between optimism and caution was captured in one nurse’s statement:

While I believe AI will enhance our efficiency, we must ensure it complements rather than compromises the human aspect of nursing care (Participant 33, Female, 45 years).

Participants frequently highlighted the necessity of ensuring that AI integration aligns with nursing's core principles, such as empathy, compassion, and individualised patient care. This cautious approach reflects a commitment to preserving nursing's holistic nature while embracing technological advancements.

Subtheme 4b: expectations for gradual implementation

55% of participants emphasised the importance of a phased or gradual approach to AI adoption in nursing practice. This perspective was rooted in the belief that a slower, deliberate implementation process would allow nurses sufficient time to adapt to new systems, understand their functionality, and gain confidence in their use. Participants noted that abrupt or poorly planned rollouts could lead to confusion, resistance, and suboptimal use of AI tools. One participant explained:

We need time to adapt and learn how to best use these tools in our daily practice (Participant 9, Male, 38 years).

Nurses stressed that gradual integration would also provide opportunities for iterative improvements based on feedback, ensuring that the tools align with clinical workflows and address the needs of both staff and patients. This phased approach was seen as essential for fostering acceptance and long-term success in AI adoption, particularly in environments where familiarity with such technologies is limited.

Overview of key themes

The study identified four interconnected domains influencing AI adoption in nursing practice: understanding, facilitators, barriers, and attitudes. These themes and subthemes are summarised in Table 2 and visually represented in Fig. 3, which highlights the relationships among these domains. Participants’ understanding of AI varied widely. While 65% (n = 31) demonstrated basic comprehension of AI’s role in automating tasks and supporting decision-making, 25% (n = 12) held misconceptions, often fearing it might replace human roles. A smaller group (10%, n = 5) exhibited advanced understanding, informed by direct experience with AI tools. Facilitators of adoption were led by perceived benefits to patient care, noted by 85% (n = 41), including improved diagnostic accuracy and early interventions. Organisational support (70%, n = 34), comprehensive training (75%, n = 36), and user-friendly systems were also crucial for integration. Barriers included technical challenges (60%, n = 29), ethical concerns about data privacy (55%, n = 26), and fears of job displacement or deskilling (45%, n = 22). Scepticism about AI’s reliability, reported by 40% (n = 19), further emphasised the need for evidence-based implementation. Attitudes toward AI reflected cautious optimism (70%, n = 34), with participants stressing the importance of preserving the human aspect of nursing. A phased approach to implementation was preferred by 55% (n = 26) to allow for adaptation and alignment with workflows. Figure 3 illustrates these findings, showing how understanding AI underpins facilitators and barriers, which collectively shape attitudes toward its adoption in nursing. This integrated framework reflects the nuanced perspectives of participants, balancing optimism with practical concern.

Discussion

This qualitative investigation provides significant insights into the complex dynamics of artificial intelligence (AI) adoption in nursing practice, revealing multifaceted perspectives that both challenge and extend the current understanding of technology integration in healthcare settings. The study's findings, derived from a diverse sample of 48 registered nurses, illuminate the intricate interplay between professional experience, organisational context, and technological readiness that shapes AI adoption patterns. The demographic composition is predominantly female nurses (77.1%) with bachelor's degrees (66.7%) and an average of 11.3 years of experience, which reflects typical nursing workforce characteristics and provides a robust foundation for understanding technology acceptance across different professional and educational levels [95–97].

The variation in AI understanding across specialities emerged as a crucial finding with significant implications for implementation strategies. Critical care nurses (20.8%) demonstrated particularly sophisticated perspectives on AI's potential, especially inpatient monitoring and early intervention scenarios. For instance, one critical care nurse remarked, "AI systems can analyse patient data much faster than we can. This could help us identify deteriorating patients earlier, potentially saving lives" (Participant 23, Female, 42 years). This advanced understanding likely stems from their regular exposure to complex technological systems and high-stakes decision-making environments, aligning with Khanijahani et al.'s (2022) findings regarding speciality-specific technology adoption patterns [98]. The study revealed four deeply interconnected themes that significantly influence AI adoption: understanding, facilitators, barriers, and attitudes. The distribution of AI understanding levels basic (65%), misconceptions (25%), and advanced (10%) provides crucial insights for educational intervention design. This pattern aligns with Ronquillo et al.'s (2021) research on nursing informatics competency development [64], highlighting specific areas requiring targeted support. The misconceptions identified, particularly fears about AI replacing nursing roles, echo concerns documented in broader healthcare technology studies [99–101] but take on unique significance in nursing, where human interaction and clinical judgment are fundamental to practice.

Organisational support emerged as a critical facilitator, with 70% of participants emphasising its importance in successful AI adoption. This finding extends beyond traditional Technology Acceptance Model (TAM) constructs, suggesting the need for a more comprehensive theoretical framework that acknowledges the complex organisational dynamics in healthcare settings. One nurse manager noted, When our hospital leadership actively promotes and invests in AI, it creates a culture of innovation. This makes staff more open to embracing these tools (Participant 8, Female, 47 years). This observation aligns with Seibert et al.'s (2023) research on organisational readiness for healthcare innovation [102], highlighting nursing-specific implementation challenges. The technical challenges and infrastructure limitations, cited by 60% of participants, revealed significant disparities between healthcare facilities. This digital divide particularly affects smaller and rural hospitals, potentially exacerbating existing healthcare inequities. Participants from smaller facilities emphasised the need for substantial infrastructure upgrades. One nurse expressed, Our current IT system is outdated. I worry that implementing AI would require a complete overhaul (Participant 5, Male, 39 years). These findings support Walter's (2024) research on technological disparities in healthcare settings [97], adding crucial insights about their impact on nursing practice.

The ethical dimensions of AI adoption, raised by 55% of participants, extended beyond basic privacy concerns to encompass fundamental questions about patient care quality and professional autonomy. Nurses expressed particular concern about AI's impact on the nurse-patient relationship. For example, a senior nurse remarked, While I see the benefits of AI, I am concerned about patient privacy. How can we ensure that sensitive data isn't misused?"(Participant 14, Female, 52 years). These ethical considerations align with recent literature on healthcare AI ethics [24, 69, 103–106], highlighting nursing-specific concerns about maintaining human connection in increasingly technological care environments. A notable generational divide emerged regarding perceptions of AI's impact on nursing practice. While experienced nurses expressed concerns about skill erosion and job displacement (45% of participants), younger practitioners viewed AI as an opportunity for role enhancement. This dichotomy reflects broader patterns in healthcare technology adoption [107–109] and suggests the need for age-specific implementation strategies that address both groups' concerns while leveraging their different perspectives.

The development of the Technology Acceptance Model for AI in Nursing (TAM-AIN) represents a significant theoretical contribution to the field. This enhanced framework incorporates four crucial dimensions emerging from our data: ethical alignment, organisational readiness, professional identity preservation, and technical infrastructure capacity. TAM-AIN extends beyond traditional technology acceptance models by specifically addressing nursing practice's unique contexts and challenges [110], providing a more comprehensive framework for understanding and facilitating AI adoption in healthcare settings. The role of education and training emerged as crucial, with 75% of participants emphasising their importance. Participants consistently highlighted the need for comprehensive training programs that address both technical competencies and ethical considerations. One nurse shared, The training we received on our new AI-assisted fetal monitoring system was excellent. It made me feel confident in using the technology and understanding its benefits (Participant 42; Female, 36 years). This finding suggests the need for integrated approaches to professional development that combine technical skills with critical thinking about AI's role in nursing practice.

Implications for research and practice

These findings have significant implications for both research and practice. Healthcare organisations must develop comprehensive implementation strategies that address the varied needs and concerns identified across different nursing specialities and experience levels. Specific recommendations include:

Firstly, it is essential to develop speciality-specific AI training programs that address technical skills and ethical considerations. Tailoring training to the unique requirements of each nursing speciality can enhance competency and confidence in using AI tools effectively. Secondly, establishing robust organisational support structures is crucial. This includes having AI champions within the organisation and providing ongoing professional development opportunities to foster a culture of innovation. Leadership commitment can significantly influence staff openness to embrace AI technologies. Thirdly, implementing clear ethical frameworks and data governance policies is necessary to address privacy concerns and build trust among staff and patients. Transparent policies can alleviate fears regarding data misuse and ensure compliance with legal and ethical standards. Fourthly, creating mentorship programs that pair technologically experienced nurses with those requiring additional support can facilitate knowledge transfer and confidence building. Such programs can help bridge the gap between different experience levels and promote a collaborative learning environment.

Lastly, investment in infrastructure development is imperative to address technical barriers and ensure equitable access to AI technologies across different healthcare settings [95, 102]. Upgrading IT systems and providing the necessary resources can mitigate disparities between facilities. Future research directions should include longitudinal studies examining the evolution of nurses' attitudes and practices as AI becomes more prevalent in healthcare. Cross-cultural comparative research can explore variations in healthcare system approaches to AI adoption in different cultural and organisational contexts. Additionally, intervention studies testing the effectiveness of different implementation strategies and educational interventions in promoting AI adoption are recommended. Investigating patient outcomes in AI-enhanced nursing environments and developing AI-specific nursing competency frameworks are also valuable areas for further study.

Strengths and limitations

This study's strengths lie in its methodological rigour, diverse sample, and theoretical grounding. The focus group methodology enabled a rich exploration of perspectives while maintaining systematic analysis procedures. However, several limitations warrant consideration. The geographic focus on Al-Kharj, Saudi Arabia, may reflect specific cultural and organisational contexts that differ from other settings. The rapid modernisation and substantial investment in healthcare technology in Saudi Arabia may have influenced participants' perspectives differently from other global contexts. Additionally, the study's cross-sectional nature captures attitudes at a specific moment in an evolving technological landscape. Future research should address these limitations through longitudinal designs and more diverse geographic sampling to capture the dynamic nature of AI adoption in nursing practice.

Conclusion

This study offers a comprehensive understanding of the factors influencing AI adoption in nursing practice, highlighting the need to balance technological advancements with the preservation of the essential human elements of care. Successful implementation of AI hinges on addressing technical challenges, providing adequate education and training, ensuring ethical use and data privacy, and fostering organisational support. The development of the Technology Acceptance Model for AI in Nursing (TAM-AIN) extends existing models by incorporating dimensions specific to nursing, such as ethical alignment, organisational readiness, professional identity preservation, and technical infrastructure capacity. This model provides a robust foundation for future research and practical application in healthcare settings. Nurses in this study expressed cautious optimism about AI, recognising its potential to enhance patient care while voicing concerns about maintaining empathy and interpersonal relationships. Tailored implementation strategies, including speciality-specific training programs and strong leadership commitment, are crucial to addressing these concerns and facilitating acceptance. As healthcare continues to evolve with technological innovations, it is imperative to ensure that AI integration enhances rather than diminishes the core values of nursing. By prioritising nurse-centred approaches and addressing identified barriers, healthcare organisations and policymakers can effectively adopt AI technologies, ultimately improving patient care and outcomes.

Authors’ contributions

M.M.A. and O.M.E.R. contributed equally to this work as co-first authors. They conceptualized the study, developed the methodology, conducted the literature search, and wrote the original draft of the manuscript. A.N.A. assisted with data extraction, conducted the quality assessment of included studies, and critically reviewed the manuscript. M.G.E. contributed to the data analysis, interpretation of results, and provided critical insights for the discussion. S.A. supervised the project, contributed to the study design, and reviewed and edited the final manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This research is funded by the Deanship of Graduate Studies and Scientific Research at Jouf University through the Fast-Track Research Funding Program.

Data availability

Data will be available upon request.

Declarations

Ethics approval and consent to participate

This study was approved by the Standing Committee of Bioethics Research at Prince Sattam bin Abdulaziz University (Approval No. SCBR-309/2024). Informed consent was obtained from all participants involved in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.