Assessment of Knowledge and Attitudes of Healthcare Personnel Towards Artificial Intelligence Technologies in Greece: A Survey Study

Abstract

Artificial intelligence (AI) is progressively being utilized in the healthcare sector to enhance efficiency, alleviate administrative burdens, and improve patient care outcomes. In the secondary healthcare sector, AI presents a range of opportunities as well as challenges. This study investigates the viewpoints of healthcare professionals regarding the adoption of AI in Greece, emphasizing the anticipated advantages and apprehensions associated with its integration. A cross-sectional descriptive study was carried out to collect responses from healthcare professionals at the General Hospital of “Evangelismos”, which is the largest hospital in Athens, Greece. A questionnaire was utilized and distributed over a period of four months, involving 513 registered healthcare professionals (comprising 136 physicians, 235 nursing staff, and 142 other healthcare personnel). Each participant had a minimum of one year of clinical experience and was selected using a convenience sampling method. The questionnaire comprised two parts: one focused on evaluating the AI knowledge and attitudes of healthcare professionals, and the other collected demographic data. The overall comprehension of knowledge pertaining to AI among healthcare professionals was evaluated as moderate, resulting in a mean score of 3.39. A distinction exists among different personnel categories, with physicians (M = 3.73) demonstrating a greater understanding of AI and a firm conviction that AI cannot supplant human positions. Conversely, nursing personnel appear to express apprehension regarding the implications of AI on the human experience, with a notable concern about potential replacement and job loss (M = 2.63), which was identified as the lowest-ranked issue. This latter concern is also echoed by other healthcare personnel (M = 2.90). Nevertheless, the majority of participants regard the prospective use of AI favorably, demonstrate confidence in its application, and contend that the benefits outweigh the possible risks. Sufficient training and ongoing updates would enhance employees’ comprehension of AI and their awareness of its potential benefits within the healthcare sector.

1. Introduction

Global health systems encounter considerable challenges in achieving their primary healthcare objectives, which encompass enhancing population health, improving the patient care experience, boosting caregivers’ satisfaction, and controlling escalating healthcare expenses. Environmental factors, including air pollution and severe weather occurrences, have notably exacerbated the burden of disease and mortality, particularly among vulnerable populations. Additionally, the incidence of chronic illnesses and conditions, including heart disease, diabetes, cancer, Alzheimer’s, and obesity, has been progressively increasing. These ailments often lead to complications that necessitate ongoing treatment, potentially resulting in the requirement for assisted living arrangements or even long-term care. This situation consequently presents challenges related to patient discharge and bed overcrowding. The management of chronic illnesses typically requires prolonged medication and modifications to lifestyle, with certain conditions limiting daily activities and necessitating additional support services. Moreover, this demographic is also more prone to mental health challenges, which further complicates and exacerbates the resource demands of healthcare delivery [1,2,3,4,5].

The surge in healthcare expenditure is frequently attributed to routine consultations, outpatient services, prescriptions, and emergency interventions. In this regard, workforce burnout has emerged as a significant issue within health systems, affecting professionals across various fields, including physicians, nursing staff, technicians, and administrative personnel, who suffer from chronic physical, emotional, and mental exhaustion due to prolonged and excessive stress. The ongoing increase in healthcare expenses is influenced by numerous external elements, such as cultural norms, governmental policies, inflationary pressures, and a demographic shift towards an older population. These factors frequently impose financial limitations on hospitals, clinics, and primary care facilities, thereby restricting their capacity to allocate funds towards vital resources, equipment, and infrastructure. As populations expand and individuals experience extended life spans, their healthcare requirements also escalate, with the nature of care needed often demanding greater resources, which in turn drives up demand and costs. The COVID-19 pandemic has burdened healthcare systems with a significant backlog of care, compounded by new patients who delayed seeking assistance during the pandemic, leading to more intricate health issues [6,7,8].

The enhancement of Electronic Health Record (EHR) implementation, advanced diagnostic tools, and the adoption of telemedicine has greatly improved the efficiency of healthcare services, while also establishing a foundation for the smooth integration of artificial intelligence (AI) into contemporary medical practices. The presence of extensive data sets and the advancement of machine learning algorithms have facilitated the training of AI models, which, when paired with improvements in computing power, have led to the proliferation of AI applications within the healthcare sector. Machine learning, a branch of AI, has been particularly significant, enabling computers to analyze and learn from data patterns, thereby improving their capacity to make predictions and decisions. The application of AI in medical imaging, disease diagnosis, and treatment optimization has showcased its potential to empower healthcare professionals and enhance patient outcomes. As time progresses, AI technologies represent the forthcoming wave of digital innovation influencing healthcare settings and are essential for tackling numerous challenges, while also offering distinct advantages in areas such as diagnostic support, patient monitoring, personalized healthcare, and administrative tasks [9,10,11,12].

Despite the simultaneous advancements in technology, a significant knowledge gap persists concerning healthcare providers’ views on AI in healthcare. Healthcare professionals, including physicians, nurses, radiologists, and various other medical staff, have historically played a vital role in the healthcare ecosystem. Their expertise, compassion, and dedication have been instrumental in delivering medical care, safeguarding patient well-being, and fostering trust between patients and healthcare providers. These individuals have invested years in education and practice, honing specialized skills and knowledge that empower them to provide medical and nursing services that include diagnosis, treatment, rehabilitation, and emotional support. Their duties extend beyond the mere application of medical knowledge; they also encompass the establishment of relationships with patients, understanding their unique situations, and making compassionate decisions that go beyond the strictly clinical aspects of healthcare. In terms of this, the effective integration of AI in clinical and nursing settings relies on the awareness, acceptance, and readiness of healthcare providers to embrace these technologies. Numerous researchers have shown that an increased understanding of AI among healthcare professionals is associated with more favorable attitudes towards its application. Research indicates that insufficient knowledge, apprehensions regarding data privacy, ethical dilemmas, and a deficit of trust in the transparency of algorithms remain significant barriers to the adoption of AI in healthcare environments. Nevertheless, earlier research predominantly concentrated on the effectiveness of AI technologies in the healthcare sector, overlooking the factors related to healthcare professionals that are crucial for their smooth incorporation into current workflows [13,14,15]. Simultaneously, there is apprehension that in the absence of collaboration and support from healthcare professionals, the complete potential of AI may remain unfulfilled or may not be implemented in clinical practice, resulting in a significant opportunity being overlooked. Given the constraints surrounding the investigation into workforce readiness and the preparation for this emerging generation of technology, this research aims to illuminate healthcare professionals’ comprehension of AI concepts and their viewpoints regarding its implementation. By concentrating on these aspects, the study aims to uncover the extent of their knowledge regarding AI applications and evaluate their attitudes towards its adoption. Comprehending the roles and viewpoints of healthcare professionals in the integration of AΙ is a focal point of worldwide attention and indicates the development of strategies that promote a synergistic partnership between human skills and technological advancements [16,17,18,19,20].

Consequently, the primary aim of this study was to assess the existing knowledge of healthcare personnel and evaluate their perceptions of AI technologies within hospital environments, as well as the potential influence on their professional practices. The second specific goal was to explore the influence of sociodemographic factors on the adoption of AI. The third objective was to examine the viewpoints of personnel concerning the ethical dilemmas linked to the progress of artificial intelligence and to pinpoint deficiencies in their training. The final objective was to suggest actionable strategies for healthcare organizations to adopt AI as an auxiliary tool, while preserving the human aspect of healthcare.

2. Methods

2.1. Instrument

This cross-sectional study is organized into two sections. Part one, consisting of 15 items, assessed healthcare professionals’ perceptions of AI, without taking into account the different ways in which the various groups become aware of AI (such as scientific papers, coursework, or lectures, etc.). This assessment utilized a questionnaire adopted in 2020 from Abdullah & Fakieh [21]. This part was segmented into three subparts: questions 1–4 pertained to employees’ knowledge of AI, questions 5–14 discussed the advantages and challenges associated with AI, while question 15 examined employees’ views on the future of AI in the healthcare sector. Responses were gathered using a 5-point Likert scale, where 1 represents Strongly Agree, 2 signifies Agree, 3 indicates Neutral, 4 denotes Disagree, and 5 stands for Strongly Disagree. The items were formulated in both positive and negative directions, necessitating that approximately half of them be reverse scored. The scoring system was based on the average of the scores. The professionals’ perceptions were categorized as low (1.0–2.60), moderate (2.61–3.40), and high (3.41–5.0). The second part encompassed sociodemographic information, including participants’ age, gender, marital status, educational background, and work-related details such as professional category and duration of experience. All questions were presented in the Greek language, following their translation from English, and were readily comprehensible, as there were no unusual expressions or idiomatic phrases.

2.2. Settings and Participants

The survey was conducted between September 2024 and May 2025 at the General Hospital of Evangelismos in Athens, the capital of Greece. In the pursuit of refining the questionnaire, a pilot testing phase was carried out, aiming to identify any potential issues with 20 healthcare providers: five physicians (25%), ten nursing personnel (50%), and five other healthcare professionals (25%). As this distribution reflects the healthcare workforce composition in Greek metropolitan hospitals, no modifications were necessary. The reliability of the pilot study was assessed, revealing a Cronbach’s alpha of 0.75, which confirmed the reliability of the instrument. This pretesting phase enhanced the instrument’s validity by confirming that the items were semantic, conceptually equivalent, and comprehensively understood by the intended population.

The researchers gathered the data through a paper-based survey, and an adequate sample size was achieved through convenience and snowball sampling. Convenience and snowball sampling play a vital role for researchers by offering efficient and economical means to engage with hard-to-reach or specialized populations when conventional methods are inadequate. These approaches provide a flexible solution for swiftly collecting preliminary data and accessing groups with limited numbers or distinct characteristics [22,23]. The inclusion criteria were as follows: (a) healthcare providers, including physicians, nursing personnel, and other healthcare professionals (such as laboratory technologists, therapists, and radiologists, etc.) who were employed in the specific hospital and possessed valid qualification certificates, and (b) voluntary participation in the study. For those professionals who consented to participate, an envelope containing the instruments and the consent form was provided. Consequently, the participants completed a personal and professional characterization form. Out of the 600 questionnaires distributed to 150 physicians (25%), 300 nursing personnel (50%), and 150 other healthcare professionals (25%), a significant number of 513 (85.5%) were returned: 136 physicians (26.51%), 235 nursing personnel (45.81%), and 142 other healthcare professionals (27.68%). It is noteworthy that a significant number of questionnaires were not returned by nursing personnel (65 or 21.7% of those distributed in this job category).

Respondents were assured that the study results would be utilized solely for scientific purposes. The sample size calculation was guided by Cochran’s sample size formula [24]:

$$\mathrm{N}={\displaystyle \frac{{\mathrm{Z}}^2\times \mathrm{P}\times \left(1-\mathrm{P}\right)}{{\mathsf{{\rm E}}}^2}}$$

where

• N = required sample size;
• Z = Z-score (1.96 for a 95 % confidence level);
• P = estimated proportion (0.5 for maximum variability);
• E = margin of error (0.05).

For our present study, substituting the values, we have the following:

$$\mathrm{N}={\displaystyle \frac{{\left(1.96\right)}^2\times 0.5\times \left(1-0.5\right)}{{\left(0.05\right)}^2}}=384.16 \approx 384$$

According to this calculation, it was essential to incorporate 384 healthcare professionals into the study to achieve a 95% confidence level with a 5% margin of error. Adhering to a meticulously established sample size ensures the study’s ability to detect meaningful patterns or trends within the population of healthcare workers, thereby enhancing the overall statistical power and generalizability of the findings. This approach is particularly crucial when investigating intricate topics such as the integration of AI in the healthcare sector.

2.3. Statistical Analysis

Descriptive statistics, including mean scores, standard deviations (SD), and qualitative data represented as absolute and relative frequencies, were computed for each question. In addition to descriptive statistics, this study employed inferential statistics to uncover more complex relationships within the data. The Kolmogorov–Smirnov and Shapiro–Wilk tests were conducted to assess normality. Furthermore, Spearman’s rho test was applied to investigate the associations between demographic variables and healthcare personnel’s knowledge and attitudes regarding the adoption of artificial intelligence (AI). Multiple regression assumptions (normality, homoscedasticity, and multicollinearity) were validated using Shapiro–Wilk tests for the interpretation of the AI-related worries. Reliability analysis was conducted using Cronbach’s alpha to evaluate internal consistency. The threshold for statistical significance was established at 0.05. All statistical analyses were executed using SPSS, version 30.0 (IBM Corp., Armonk, NY, USA).

2.4. Validity and Reliability Analysis

The Kolmogorov–Smirnov and Shapiro–Wilk tests for normality were employed to assess the distribution of the data. The findings indicated that the data did not follow a normal distribution, as the p-value for both tests was below 0.05. In terms of reliability, Cronbach’s alpha coefficient was calculated for the survey instrument, resulting in an overall value of α = 0.84. A commonly accepted guideline suggests that an α value of 0.80 or higher represents a very good level of reliability [25].

2.5. Ethical Considerations

The Committee of the University of Piraeus granted approval for the study protocol in February 2024 (confirmation: 20241411-17 June 2024). Additionally, the research was carried out following a thorough review and the acquisition of written approvals from the pertinent institutional ethics and research committee at the General Hospital of Evangelismos in Athens, Greece (approval number: 326-19 June 2024). Furthermore, this research constituted a perception survey that did not involve any medical interventions or the gathering of patient medical records; no personally identifiable or medical information was obtained and the participants were healthcare professionals instead of patients. Lastly, the researcher provided each participant with information regarding the study’s purpose, and participation was voluntary, based on written informed consent obtained prior to the commencement of data collection. The anonymity of participants and the confidentiality of the data were guaranteed. The study adhered to the principles outlined in the Helsinki Declaration of 2013.

3. Results

3.1. Percentage Distribution of Healthcare Professionals’ Characteristics

Regarding the professionals’ roles, 26.5% were physicians, 45.8% were nursing personnel, and 27.7% were other healthcare professionals. There was clear female dominance in the group, with females constituting approximately 79%, while males made up the remaining 21%. The age group of 46–55 years was the most represented (39.4%), followed by the 36–45 years age group (22.6%). The 26 to 35 years age group was 21.0%, and over 56 years made up approximately 17.0%. A considerable proportion of the participants were married, accounting for 67.3%, whereas 17.1% were single, 11.7% identified as divorced, and merely 3.9% were widowed. A significant proportion had a master’s degree or doctorate (36.8%), 34.9% held a bachelor of science degree, and a significant proportion had a diploma (28.3%). Regarding their position in hospital, the vast majority of respondents were employees (97.7%), and only a few were supervisors (1.5%) or held a director position (0.8%). As for their employment status, the majority were permanent employees (76.6%), while 120 individuals (23.4%) were temporary personnel. In terms of their experience in the field, 19.1% possessed less than 5 years, 16% had between 6–10 years, 10.9% had 11–15 years, and 14% had 16–20 years, whereas the majority (40%) had over 20 years of experience. About 50.3% of healthcare professionals stated that they had great financial difficulties and 34.1% asserted that they were able to meet their financial responsibilities, albeit with little money remaining afterwards. In contrast, 12.5% faced significant financial challenges, while only 3.1% experienced financial comfort.

3.2. Healthcare Professionals’ Perceptions of Artificial Intelligence

Research within the field of psychology has demonstrated that attitudes serve as significant indicators of behavior. Consequently, examining general attitudes towards AI technologies in healthcare is crucial for comprehending professionals’ perspectives and formulating effective policies related to AI. Previous studies indicate a range of attitudes towards AI.

Table 1. Healthcare personnel’s perceptions of artificial intelligence technologies.

		Category of Healthcare Personnel		Strongly Agree	Agree	Total Agreement	Neutrality	Disagree	Strongly Disagree	Total Disagreement
Item	Question			1	2	= 1 + 2	4	5	6	= 5 + 6
Q1	I have good knowledge of AI	Physicians	(Ν)	18	81	99	20	15	2	17
			(%)	13.24%	59.56%	72.79%	14.71%	11.03%	1.47%	12.50%
		Nursing	(Ν)	41	135	176	54	5	0	5
		personnel	(%)	17.45%	57.45%	74.89%	22.98%	2.13%	0.00%	2.13%
		Other healthcare	(Ν)	23	68	91	48	3	0	3
		personnel	(%)	16.20%	47.89%	64.08%	33.80%	2.11%	0.00%	2.11%
		Total personnel	(Ν)	82	284	366	122	23	2	25
			(%)	15.98%	55.36%	71.35%	23.78%	4.48%	0.39%	4.87%
Q2	AI abilities are superior to human experience	Physicians	(Ν)	5	23	28	28	71	9	80
			(%)	3.68%	16.91%	20.59%	20.59%	52.21%	6.62%	58.82%
		Nursing	(Ν)	23	66	89	41	80	25	105
		personnel	(%)	9.79%	28.09%	37.87%	17.45%	34.04%	10.64%	44.68%
		Other healthcare	(Ν)	17	14	31	24	61	26	87
		personnel	(%)	11.97%	9.86%	21.83%	16.90%	42.96%	18.31%	61.27%
		Total personnel	(Ν)	45	103	148	93	212	60	272
			(%)	8.77%	20.08%	28.85%	18.13%	41.33%	11.70%	53.02%
Q3	AI could replace me in my job	Physicians	(Ν)	0	5	5	16	75	40	115
			(%)	0.00%	3.68%	3.68%	11.76%	55.15%	29.41%	84.56%
		Nursing	(Ν)	29	106	135	25	74	1	75
		personnel	(%)	12.34%	45.11%	57.45%	10.64%	31.49%	0.43%	31.91%
		Other healthcare	(Ν)	11	60	71	16	42	13	55
		personnel	(%)	7.75%	42.25%	50.00%	11.27%	29.58%	9.15%	38.73%
		Total personnel	(Ν)	40	171	211	57	191	54	245
			(%)	7.80%	33.33%	41.13%	11.11%	37.23%	10.53%	47.76%
Q4	I have high hopes about AI applications in the healthcare sector	Physicians	(Ν)	30	61	91	18	26	1	27
			(%)	22.06%	44.85%	66.91%	13.24%	19.12%	0.74%	19.85%
		Nursing	(Ν)	72	104	176	45	12	2	14
		personnel	(%)	30.64%	44.26%	74.89%	19.15%	5.11%	0.85%	5.96%
		Other healthcare	(Ν)	23	82	105	24	12	1	13
		personnel	(%)	16.20%	57.75%	73.94%	16.90%	8.45%	0.70%	9.15%
		Total personnel	(Ν)	125	247	372	87	50	4	54
			(%)	24.37%	48.15%	72.51%	16.96%	9.75%	0.78%	10.53%
Q5	AI can speed up the process in healthcare	Physicians	(Ν)	40	90	130	6	0	0	0
			(%)	29.41%	66.18%	95.59%	4.41%	0.00%	0.00%	0.00%
		Nursing	(Ν)	57	157	214	21	0	0	0
		personnel	(%)	24.26%	66.81%	91.06%	8.94%	0.00%	0.00%	0.00%
		Other healthcare	(Ν)	29	85	114	27	1	0	1
		personnel	(%)	20.42%	59.86%	80.28%	19.01%	0.70%	0.00%	0.70%
		Total personnel	(Ν)	126	332	458	54	1	0	1
			(%)	24.56%	64.72%	89.28%	10.53%	0.19%	0.00%	0.19%
Q6	AI can help reduce the number of medical errors	Physicians	(Ν)	46	76	122	14	0	0	0
			(%)	33.82%	55.88%	89.71%	10.29%	0.00%	0.00%	0.00%
		Nursing	(Ν)	75	134	209	26	0	0	0
		personnel	(%)	31.91%	57.02%	88.94%	11.06%	0.00%	0.00%	0.00%
		Other healthcare	(Ν)	53	63	116	26	0	0	0
		personnel	(%)	37.32%	44.37%	81.69%	18.31%	0.00%	0.00%	0.00%
		Total personnel	(Ν)	174	273	447	66	0	0	0
			(%)	33.92%	53.22%	87.13%	12.87%	0.00%	0.00%	0.00%
Q7	AI can deliver clinically relevant, vast amounts of high-quality data in real time	Physicians	(Ν)	22	85	107	28	1	0	1
			(%)	16.18%	62.50%	78.68%	20.59%	0.74%	0.00%	0.74%
		Nursing	(Ν)	74	137	211	11	13	0	13
		personnel	(%)	31.49%	58.30%	89.79%	4.68%	5.53%	0.00%	5.53%
		Other healthcare	(Ν)	42	72	114	19	9	0	9
		personnel	(%)	29.58%	50.70%	80.28%	13.38%	6.34%	0.00%	6.34%
		Total personnel	(Ν)	138	294	432	58	23	0	23
			(%)	26.90%	57.31%	84.21%	11.31%	4.48%	0.00%	4.48%
Q8	AI has no space–time constraint	Physicians	(Ν)	17	93	110	16	10	0	10
			(%)	12.50%	68.38%	80.88%	11.76%	7.35%	0.00%	7.35%
		Nursing	(Ν)	5	175	180	55	0	0	0
		personnel	(%)	2.13%	74.47%	76.60%	23.40%	0.00%	0.00%	0.00%
		Other healthcare	(Ν)	6	94	100	39	3	0	3
		personnel	(%)	4.23%	66.20%	70.42%	27.46%	2.11%	0.00%	2.11%
		Total personnel	(Ν)	28	362	390	110	13	0	13
			(%)	5.46%	70.57%	76.02%	21.44%	2.53%	0.00%	2.53%
Q9	AI has no emotional exhaustion or physical limitation	Physicians	(Ν)	48	82	130	4	2	0	2
			(%)	35.29%	60.29%	95.59%	2.94%	1.47%	0.00%	1.47%
		Nursing	(Ν)	77	157	234	1	0	0	0
		personnel	(%)	32.77%	66.81%	99.57%	0.43%	0.00%	0.00%	0.00%
		Other healthcare	(Ν)	45	97	142	0	0	0	0
		personnel	(%)	31.69%	68.31%	100.00%	0.00%	0.00%	0.00%	0.00%
		Total personnel	(Ν)	170	336	506	5	2	0	2
			(%)	33.14%	65.50%	98.64%	0.97%	0.39%	0.00%	0.39%
Q10	AI cannot be used to provide opinions in unexpected situations	Physicians	(Ν)	18	70	88	16	26	6	32
			(%)	13.24%	51.47%	64.71%	11.76%	19.12%	4.41%	23.53%
		Nursing	(Ν)	52	177	229	4	1	1	2
		personnel	(%)	22.13%	75.32%	97.45%	1.70%	0.43%	0.43%	0.85%
		Other healthcare	(Ν)	13	116	129	6	4	3	7
		personnel	(%)	9.15%	81.69%	90.85%	4.23%	2.82%	2.11%	4.93%
		Total personnel	(Ν)	83	363	446	26	31	10	41
			(%)	16.18%	70.76%	86.94%	5.07%	6.04%	1.95%	7.99%
Q11	AI is not flexible enough to be applied to every patient	Physicians	(Ν)	17	42	59	20	37	20	57
			(%)	12.50%	30.88%	43.38%	14.71%	27.21%	14.71%	41.91%
		Nursing	(Ν)	60	174	234	0	1	0	1
		personnel	(%)	25.53%	74.04%	99.57%	0.00%	0.43%	0.00%	0.43%
		Other healthcare	(Ν)	37	88	125	7	8	2	10
		personnel	(%)	26.06%	61.97%	88.03%	4.93%	5.63%	1.41%	7.04%
		Total personnel	(Ν)	114	304	418	27	46	22	68
			(%)	22.22%	59.26%	81.48%	5.26%	8.97%	4.29%	13.26%
Q12	AI is difficult to apply to controversial subjects	Physicians	(Ν)	23	83	106	8	20	2	22
			(%)	16.91%	61.03%	77.94%	5.88%	14.71%	1.47%	16.18%
		Nursing	(Ν)	102	129	231	4	0	0	0
		personnel	(%)	43.40%	54.89%	98.30%	1.70%	0.00%	0.00%	0.00%
		Other healthcare	(Ν)	53	74	127	11	4	0	4
		personnel	(%)	37.32%	52.11%	89.44%	7.75%	2.82%	0.00%	2.82%
		Total personnel	(Ν)	178	286	464	23	24	2	26
			(%)	34.70%	55.75%	90.45%	4.48%	4.68%	0.39%	5.07%
Q13	AI has a low ability to sympathize and consider the emotional well-being of the patient	Physicians	(Ν)	50	65	115	21	0	0	0
			(%)	36.76%	47.79%	84.56%	15.44%	0.00%	0.00%	0.00%
		Nursing	(Ν)	55	180	235	0	0	0	0
		personnel	(%)	23.40%	76.60%	100.00%	0.00%	0.00%	0.00%	0.00%
		Other healthcare	(Ν)	36	101	137	5	0	0	0
		personnel	(%)	25.35%	71.13%	96.48%	3.52%	0.00%	0.00%	0.00%
		Total personnel	(Ν)	141	346	487	26	0	0	0
			(%)	27.49%	67.45%	94.93%	5.07%	0.00%	0.00%	0.00%
Q14	AI was developed by a specialist with little clinical experience in medical practice	Physicians	(Ν)	0	0	0	59	52	25	77
			(%)	0.00%	0.00%	0.00%	43.38%	38.24%	18.38%	56.62%
		Nursing	(Ν)	0	59	59	7	108	61	169
		personnel	(%)	0.00%	25.11%	25.11%	2.98%	45.96%	25.96%	71.91%
		Other healthcare	(Ν)	0	40	40	28	49	25	74
		personnel	(%)	0.00%	28.17%	28.17%	19.72%	34.51%	17.61%	52.11%
		Total personnel	(Ν)	0	99	99	94	209	111	320
			(%)	0.00%	19.30%	19.30%	18.32%	40.74%	21.64%	62.38%
Q15	Do you believe that artificial intelligence will contribute to the development of the Health Sector?	Physicians	(Ν)	31	101	132	4	0	0	0
			(%)	22.79%	74.26%	97.06%	2.94%	0.00%	0.00%	0.00%
		Nursing	(Ν)	76	153	229	6	0	0	0
		personnel	(%)	32.34%	65.11%	97.45%	2.55%	0.00%	0.00%	0.00%
		Other healthcare	(Ν)	28	97	125	17	0	0	0
		personnel	(%)	19.72%	68.31%	88.03%	11.97%	0.00%	0.00%	0.00%
		Total personnel	(Ν)	135	351	486	27	0	0	0
			(%)	26.32%	68.42%	94.74%	5.26%	0.00%	0.00%	0.00%

provides a detailed presentation of healthcare providers’ responses regarding AI technologies in hospital settings.

Table 2. Means and standard deviations of healthcare professionals’ perceptions of AI technologies.

		Physicians		Nursing Personnel		Other Healthcare Personnel		Total Personnel
Item	Question	M	SD	M	SD	M	SD	M	SD
Q1	I have good knowledge of AI	3.72	0.884	3.20	1.093	3.16	1.076	3.33	1.062
Q2	AI abilities are superior to human experience	3.41	0.970	3.08	1.199	3.46	1.241	3.27	1.167
Q3	AI could replace me in my job	4.10	0.743	2.63	1.068	2.90	1.181	3.09	1.197
Q4	I have high hopes about AI applications in the healthcare sector	3.68	1.045	3.99	0.884	3.80	0.836	3.85	0.924
Overall perception of knowledge about AI (Q 1–4)		3.73	0.437	3.22	0.456	3.33	0.462	3.39	0.498
Q5	AI can speed up the process in healthcare	4.25	0.527	4.15	0.557	4.00	0.652	4.14	0.584
Q6	AI can help reduce the number of medical errors	4.24	0.623	4.21	0.623	4.19	0.724	4.21	0.651
Q7	AI can deliver clinically relevant, vast amounts of high-quality data in real time	3.94	0.630	4.16	0.749	4.04	0.829	4.07	0.747
Q8	AI has no space–time constraint	3.86	0.722	3.79	0.459	3.73	0.573	3.79	0.572
Q9	AI has no emotional exhaustion or physical limitation	4.29	0.598	4.32	0.478	4.32	0.467	4.31	0.509
Q10	AI cannot be used to provide opinions in unexpected situations	2.50	1.082	1.82	0.511	2.07	0.659	2.07	0.791
Q11	AI is not flexible enough to be applied to every patient	3.01	1.297	1.75	0.461	1.94	0.815	2.14	1.001
Q12	AI is difficult to apply to controversial subjects	2.23	0.950	1.58	0.528	1.76	0.714	1.80	0.759
Q13	AI has a low ability to sympathize and consider the emotional well-being of the patient	1.79	0.693	1.77	0.424	1.78	0.493	1.78	0.525
Q14	AI was developed by a specialist with little clinical experience in medical practice	3.75	0.748	3.73	1.107	3.42	1.080	3.65	1.024
Overall perception about the advantages and problems of AI (Q 5–14)		3.39	0.390	3.13	0.205	3.12	0.264	3.19	0.303
Overall perception about the future of AI (Q 15)		4.20	0.469	4.30	0.511	4.08	0.560	4.21	0.522

Notes: N = 513, physicians = 136, nursing personnel = 235, other healthcare personnel = 142, M = mean, and SD = standard deviation.

displays the mean scores (M) and standard deviations (SDs) of healthcare providers’ responses regarding AI technologies in hospital environments.

3.2.1. Healthcare Professionals’ Knowledge of Artificial Intelligence

In Table 1 and Table 2, questions 1 through 4 offer a summary of the participants’ understanding of AI adoption within the healthcare sector, with the mean values for these questions varying between 2.63 and 4.10. The general understanding of knowledge regarding AI among healthcare personnel was assessed as moderate, yielding a mean score of 3.39 ± 0.50. Statistically, 71.35% (N = 366/513) of the participants expressed agreement with the notion that they possess a good understanding of AI, while 53.02% asserted that human capabilities surpass those of AI, and 28.85% endorsed the contrary view. There is a contrast between personnel categories, as physicians, with a mean of 3.73 ± 0.44, are shown to have higher knowledge about AI (Q1 = 3.72 ± 0.88) and a strong belief that humans cannot be replaced by AI (Q3 = 4.10 ± 0.74). Nursing personnel, with a mean of 3.22 ± 0.46, seem to be concerned about the potential of AΙ in relation to the human experience (Q2 = 3.08), and skepticism about their replacement and job loss, at Q3 = 2.63, was the lowest-ranked item on the Likert scale in the overall perception of knowledge. The latter skepticism seems to be shared by the other healthcare personnel (Q3 = 2.90), who report a knowledge level regarding AI of 3.39 ± 0.50. With a deeper look into Table 1, physicians (84.56%, N = 115/136) strongly express the opinion that AI could not be replace them; on the contrary, 57.45% (N = 135/235) of nursing personnel and half of other healthcare personnel (N = 71/142) express concern about their possible replacement by AI. Finally, nearly three-quarters (72.51%, N = 372/513) of healthcare professionals have high expectations for the application of AI in healthcare.

3.2.2. Healthcare Personnel’s Perceptions of Advantages and Challenges of AI

Questions 5 to 14 in Table 2 provide a summary of the participants’ perspectives on the benefits and challenges related to AI in the healthcare sector, with mean values for these questions varying from 1.58 to 4.32. The general perception in this category of questions was found to be moderate, with an overall mean of 3.19 ± 0.30 across all healthcare personnel, as well as for the specific categories of healthcare personnel (physicians = 3.39, nursing personnel = 3.13, and other healthcare personnel = 3.12). Questions 5 to 9 focused on the potential advantages that AI can contribute to the healthcare sector. The belief that “AI has no emotional exhaustion or physical limitation” was the most widely accepted, receiving a significant level of endorsement (Μ = 4.31 ± 0.51), while high values were expressed for the beliefs that “AI can help reduce the number of medical errors”, which had a mean score of 4.21 ± 0.65, as well as the belief that “AI can speed up the process in healthcare” (Μ = 4.14 ± 0.58). Questions 10 to 14 examined the possible obstacles associated with AI in the healthcare sector. Consequently, the belief that “AI has a low ability to sympathize and consider the emotional well-being of the patient” was identified as the most significant drawback of AI (Μ = 1.78 ± 0.53). Furthermore, it is strongly believed that “AI is difficult to apply to controversial subjects” (Μ = 1.80 ± 0.76), “AI cannot be used to provide opinions in unexpected situations” (Μ = 2.07 ± 0.79), and “AI is not flexible enough to be applied to every patient” (Μ = 2.14 ± 1.00). In contrast, the belief that “AI was developed by a specialist with little clinical experience in medical practice”, which had a mean of 3.65 ± 1.02, was not endorsed by 62.38% of the respondents (N = 320/513), who contended that AI represents a more systematic advancement in the field of science.

3.2.3. Healthcare Personnel’s Perceptions of the Future of AI

Question 15 offers a summary of the participants’ perceptions of the future of AI in the healthcare sector. The healthcare workforce, in general, holds optimistic expectations, scoring 4.21 ± 0.52 concerning the future of AI in healthcare. This sentiment is reflected not only in the overall percentage of respondents (94.74%, N = 486/513), but also across various professional categories, including physicians (97.06%), nursing staff (97.45%), and other healthcare professionals (88.03%).

3.3. Healthcare Professionals’ Perceptions of Artificial Intelligence in Relation to the Sociodemographic Characteristics of Respondents

Table 3. Means and standard deviations regarding perceptions of AI in relation to the sociodemographic characteristics of respondents.

	Ν	%	Perception of Knowledge About AI		Perception of Advantages and Challenges of AI		Perception of the Future of AI
			(Q1–4)		(Q5–14)		(Q15)
			M	SD	M	SD	M	SD
Gender
Male	109	21.25%	3.42	0.536	3.27	0.315	4.15	0.541
Female	404	78.75%	3.38	0.488	3.18	0.296	4.23	0.515
Age
26–35 years	108	21.05%	3.56	0.516	3.25	0.333	4.16	0.496
36–45 years	116	22.61%	3.55	0.472	3.27	0.336	4.24	0.450
46–55 years	202	39.38%	3.25	0.491	3.14	0.277	4.25	0.574
>56 years	87	16.96%	3.27	0.399	3.14	0.241	4.14	0.510
Marital Status
Married	345	67.25%	3.38	0.503	3.19	0.296	4.23	0.540
Single	88	17.15%	3.59	0.469	3.28	0.367	4.11	0.440
Divorced	60	11.70%	3.13	0.443	3.15	0.190	4.27	0.548
Widowed	20	3.90%	3.36	0.349	3.01	0.271	4.20	0.410
Level of Education
Secondary education	338	65.89%	3.23	0.453	3.13	0.200	4.25	0.530
Bachelor	58	11.31%	3.70	0.462	3.36	0.431	4.22	0.497
Master’s or PhD	117	22.81%	3.69	0.427	3.31	0.397	4.10	0.498
Category of Personnel
Physicians	136	26.51%	3.73	0.437	3.39	0.390	4.20	0.469
Nursing personnel	235	45.81%	3.22	0.456	3.13	0.205	4.30	0.512
Other healthcare professionals	142	27.68%	3.33	0.462	3.12	0.264	4.08	0.560
Employment Position
Employee	501	97.66%	3.38	0.499	3.20	0.301	4.21	0.525
Supervisor	8	1.56%	3.56	0.372	3.15	0.400	4.13	0.354
Director	4	0.78%	3.44	0.718	3.08	0.386	4.25	0.500
Employment Status
Permanent	393	76.61%	3.34	0.491	3.16	0.279	4.20	0.514
Temporary	120	23.39%	3.55	0.490	3.30	0.350	4.23	0.546
Professional Experience
<5 years	98	19.10%	3.45	0.481	3.27	0.309	4.22	0.488
6–10 years	82	15.98%	3.63	0.460	3.33	0.352	4.22	0.472
11–15 years	56	10.92%	3.51	0.579	3.24	0.322	4.20	0.483
16–20 years	72	14.04%	3.42	0.460	3.07	0.310	4.08	0.496
>20 years	205	39.96%	3.21	0.452	3.13	0.229	4.25	0.570
Economic Situation
I cannot cope with my financial obligations	64	12.48%	3.27	0.457	3.15	0.205	4.25	0.563
I manage financially with great difficulties	258	50.29%	3.34	0.493	3.18	0.286	4.22	0.540
I manage financially but I do not have much left aside	175	34.11%	3.51	0.502	3.24	0.350	4.18	0.492
I am financially comfortable	16	3.12%	3.36	0.491	3.12	0.302	4.13	0.342
Total Healthcare Professionals	513	100%	3.39	0.498	3.19	0.303	4.21	0.522

Notes: N = 513. M = mean and SD = standard deviation.

displays the mean scores (M) along with the standard deviations (SDs) obtained across different sociodemographic categories. Additionally,

Table 4. Correlations between perceptions of artificial intelligence and sociodemographic characteristics.

		Perception of Knowledge About AI (Q1–4)	Perception of Advantages and Challenges of AI (Q5–14)	Perception of the Future of AI (Q15)	Gender	Age	Marital Status	Level of Education	Professional Category	Employment Position	Employment Status	Professional Experience	Economic Situation
Perception of knowledge about AI (Q1–4)	Correlation Coefficient	1.000	0.349 **	0.044	−0.029	−0.262 **	−0.021	0.438 **	−0.294 **	0.055	0.185 **	−0.268 **	0.170 **
	Sig. (2-tailed)	.	<0.001	0.317	0.514	<0.001	0.627	<0.001	<0.001	0.217	<0.001	<0.001	<0.001
Perception of advantages and challenges of AI (Q5–14)	Correlation Coefficient	0.349 **	1.000	0.067	−0.136 **	−0.142 **	−0.049	0.251 **	−0.244 **	−0.016	0.159 **	−0.200 **	0.056
	Sig. (2-tailed)	<0.001	.	0.132	0.002	0.001	0.264	<0.001	<0.001	0.718	<0.001	<0.001	0.204
Perception of the future of AI (Q15)	Correlation Coefficient	0.044	0.067	1.000	0.060	0.008	−0.034	−0.105 *	−0.078	−0.017	0.027	0.028	−0.059
	Sig. (2-tailed)	0.317	0.132	.	0.176	0.860	0.446	0.017	0.078	0.698	0.544	0.523	0.183
Gender	Correlation Coefficient	−0.029	−0.136 **	0.060	1.000	−0.143 **	0.046	−0.165 **	0.098 *	−0.203 **	−0.006	−0.089 *	−0.052
	Sig. (2-tailed)	0.514	0.002	0.176	.	0.001	0.299	<0.001	0.026	<0.001	0.898	0.045	0.242
Age	Correlation Coefficient	−0.262 **	−0.142 **	0.008	−0.143 **	1.000	−0.065	−0.308 **	0.410 **	0.166 **	−0.516 **	0.627 **	−0.035
	Sig. (2-tailed)	<0.001	0.001	0.860	0.001	.	0.140	<0.001	<0.001	<0.001	<0.001	<0.001	0.433
Marital Status	Correlation Coefficient	−0.021	−0.049	−0.034	0.046	−0.065	1.000	0.062	−0.073	−0.106 *	0.208 **	−0.149 **	0.084
	Sig. (2-tailed)	0.627	0.264	0.446	0.299	0.140	.	0.158	0.097	0.017	<0.001	<0.001	0.058
Level of Education	Correlation Coefficient	0.438 **	0.251 **	−0.105 *	−0.165 **	−0.308 **	0.062	1.000	−0.557 **	0.193 **	0.346 **	−0.330 **	0.410 **
	Sig. (2-tailed)	<0.001	<0.001	0.017	<0.001	<0.001	0.158	.	<0.001	<0.001	<0.001	<0.001	<0.001
Professional Category	Correlation Coefficient	−0.294 **	−0.244**	−0.078	0.098 *	0.410 **	−0.073	−0.557 **	1.000	0.014	−0.402 **	0.417 **	−0.216 **
	Sig. (2-tailed)	<0.001	<0.001	0.078	0.026	<0.001	0.097	<0.001	.	0.753	<0.001	<0.001	<0.001
Employment Position	Correlation Coefficient	0.055	−0.016	−0.017	−0.203 **	0.166 **	−0.106 *	0.193 **	0.014	1.000	−0.086	0.168 **	0.221 **
	Sig. (2-tailed)	0.217	0.718	0.698	<0.001	<0.001	0.017	<0.001	0.753	.	0.053	<0.001	<0.001
Employment Status	Correlation Coefficient	0.185 **	0.159 **	0.027	−0.006	−0.516 **	0.208 **	0.346 **	−0.402 **	−0.086	1.000	−0.622 **	0.120 **
	Sig. (2-tailed)	<0.001	<0.001	0.544	0.898	<0.001	<0.001	<0.001	<0.001	0.053	.	<0.001	0.006
Professional Experience	Correlation Coefficient	−0.268 **	−0.200 **	0.028	−0.089 *	0.627 **	−0.149 **	−0.330 **	0.417 **	0.168 **	−0.622 **	1.000	−0.123 **
	Sig. (2-tailed)	<0.001	<0.001	0.523	0.045	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001	.	0.005
Economic Situation	Correlation Coefficient	0.170 **	0.056	−0.059	−0.052	−0.035	0.084	0.410 **	−0.216 **	0.221 **	0.120 **	−0.123 **	1.000
	Sig. (2-tailed)	<0.001	0.204	0.183	0.242	0.433	0.058	<0.001	<0.001	<0.001	0.006	0.005	.

Notes: N = 513, method = Spearman’s rho, *. Correlation is significant at the 0.05 level (2-tailed) and **. Correlation is significant at the 0.01 level (2-tailed).

emphasizes notable results concerning the prevalence of adequate knowledge and favorable attitudes towards the adoption of AI in the healthcare sector.

Regarding the perception of knowledge and the advantages of AI in relation to gender, males had slightly higher mean scores compared to females; on the contrary, regarding the future of AI, females (Μ = 4.23 > 4.15) appear to have higher expectations than males. Table 4 indicates that there was no statistically significant difference in attitudes observed between males and females. Through the examination of age, it becomes evident that as individuals grow older, their understanding and views regarding the benefits of artificial intelligence tend to diminish. This trend is expected, given that younger individuals generally exhibit a stronger affinity for emerging technologies. Regarding the future of AI, individuals aged between 36 and 55 years (M = 4.24–4.25) seem to possess more optimistic expectations compared to those in the 26–35 age bracket (M = 4.16) and those who are over 56 years old (M = 4.14). Based on marital status, singles (Μ = 3.59) demonstrate superior knowledge of and attitudes towards the benefits of AI when compared to divorced individuals (Μ = 3.13) and widowed individuals (Μ = 3.01), who exhibit moderate attitudes and the lowest means in this context. Regarding future expectations, divorced individuals (Μ = 4.27) appear to have the highest aspirations. Furthermore, educational attainment significantly impacted the findings, as those holding bachelor’s, master’s, or doctorate degrees (M = 3.70) exhibited greater knowledge of and favorable attitudes towards the advantages of AI (p < 0.001). Regarding job roles, physicians (M = 3.73) distinctly display enhanced knowledge and evaluation of AI’s benefits compared to other provider categories (nursing personnel = 3.22 and other healthcare personnel = 3.33). Nevertheless, nursing personnel (Μ = 4.30) seem to possess elevated expectations regarding AI’s development, likely due to the challenges prevalent in this specific healthcare sector [21,22,23]. In terms of their hospital positions, employees (M = 3.38) show moderate knowledge relative to those in higher hospital hierarchy positions (M > 3.44). Based on their employment status, which correlates with positive attitudes (p < 0.001), temporary personnel (M = 3.55) exhibit slightly greater knowledge compared to permanent employees, who maintain moderate perceptions (M = 3.34); both groups hold moderate views on AI’s advantages and high expectations for its future. Healthcare personnel with less than 15 years of experience (M = 3.45–3.63) possess better knowledge compared to those with over 16 years of experience (M = 3.21–3.42). Lastly, individuals in a more favorable economic situation demonstrate slightly enhanced knowledge and positive attitudes (p < 0.001).

3.4. Associations of Sociodemographic Characteristics with AI Worries

In the research conducted, questions Q2, Q3, Q10, Q11, Q12, and Q13 encapsulate the concerns of healthcare professionals regarding AI, thereby forming the category of AI worries. In

Table 5. Correlations between AI worries and sociodemographic characteristics (N = 513).

	Unstandardized Coefficients		Standardized Coefficients			95.0% Confidence Interval for B		Collinearity Statistics
Model	B	Std. Error	Beta	t	p-Value	Lower Bound	Upper Bound	Tolerance	VIF
(Constant)	0.315	33.315		0.009	0.992	−65.139	65.770
Gender	0.057	0.043	0.048	1.330	0.184	−0.027	0.142	0.909	1.100
Age	0.008	0.023	0.017	0.371	0.711	−0.036	0.053	0.534	1.872
Marital Status	0.001	0.020	0.002	0.059	0.953	−0.039	0.041	0.935	1.069
Level of Education	−0.335	0.027	−0.575	−12.613	<0.001	−0.387	−0.282	0.571	1.751
Professional Category	0.013	0.030	0.020	0.445	0.657	−0.045	0.072	0.590	1.694
Employment Position	0.057	0.085	0.025	0.670	0.503	−0.110	0.224	0.853	1.172
Employment Status	−0.039	0.055	−0.034	−0.715	0.475	−0.147	0.069	0.523	1.912
Professional Experience	0.017	0.016	0.056	1.116	0.265	−0.013	0.048	0.464	2.157
Economic Situation	−0.009	0.026	−0.014	−0.354	0.723	−0.061	0.043	0.787	1.271

Notes: R = 0.636, R² = 0.404, adjusted R² = 0.392, F-value = 34.02, and p  <  0.001. Dependent variable = AI worries. Predictors = Gender, Age, Marital Status, Level of Education, Professional Category, Employment Position, Employment Status, Professional Experience, and Economic Situation. B = unstandardized coefficient, β = standardized coefficient, CI = confidence interval, and VIF = variance inflation factor (all VIF values were below 2.0, indicating no multicollinearity among predictors).

, the regression model developed accounts for approximately 40% of the variance in the AI worries variable (R² = 0.404). This indicates that the independent variables possess moderate explanatory power. Among all the sociodemographic characteristics examined, educational background has by far the strongest and most statistically significant effect (β = −0.575, p < 0.001) on AI-related concerns. As educational attainment rises, worries about AI diminish, indicating that enhanced critical thinking abilities may empower healthcare professionals to foresee implementation obstacles, rather than reflecting a general fear of technology. The other sociodemographic characteristics (Gender, Age, Marital Status, Economic Status, Employment Relationship, Position in Service, Years of Experience, and Staff Category) do not appear to significantly differentiate concern levels, suggesting that there may be a relatively homogeneous attitude across these dimensions.

4. Limitations and Future Research

This research encountered certain limitations, primarily due to the fact that only three job categories within the public hospital facilities were surveyed. Furthermore, the sample was drawn from the largest hospital in Athens, the capital city of Greece, which means that the findings may not accurately reflect the experiences of healthcare personnel in all hospitals nationwide. Future studies should encompass management-level positions and smaller healthcare facilities. Data collection was conducted solely through questionnaires, which does not eliminate the chance that participants may have responded in a manner they deemed socially acceptable, rather than reflecting their true opinions. Triangulation, incorporating interviews or alternative research methods, also addressed this limitation. Ultimately, the study employed a cross-sectional design, which restricts the ability to comprehend the progression of attitudes towards artificial intelligence over time. It is advisable that future studies incorporate data collection at various time intervals. Notwithstanding these constraints, the results of the current research add to the expanding body of literature regarding individuals’ perceptions of AI in healthcare, as it was demonstrated that these perspectives are connected to their views on AI.

5. Discussion

The incorporation of artificial intelligence in healthcare environments is progressing swiftly on an international level as hospitals across the globe investigate its capacity to improve diagnostic precision and patient involvement, optimize administrative functions, assist in clinical decision-making, and tackle shortages in the healthcare workforce, due to rapid population growth and aging [13,26,27,28,29,30,31]. This research was carried out to assess the extent of understanding concerning artificial intelligence, as well as the attitudes and acceptance related to its application, from the viewpoint of healthcare professionals in Greece. The importance of knowledge is crucial in fostering confidence and competence, especially in the use of new technologies; conversely, a deficiency in knowledge may obstruct the successful application of AI within hospital settings. Furthermore, this study is innovative in its investigation of the demographic elements affecting AI adoption, a subject that has received limited attention in the current body of literature.

Remarkably, a significant portion of respondents conveyed positive views regarding the potential of AI to enhance efficiency in healthcare environments (89.28%), particularly within the medical sector (95.6%). It is projected that AI could assist in minimizing medical errors (87.1%) by delivering clinically pertinent, extensive amounts of high-quality data in real time (84.2%). Among the participants, we inquired regarding the comparative superiority of AI over human capabilities. Less than one-third of the respondents concurred that AI would surpass human performance and only 3.7% of physicians strongly affirmed that AI is diagnostically superior. Consequently, a significant majority of physicians (84.6%) from different medical disciplines do not believe that AI will replace them. Conversely, recent investigations have indicated that image recognition technology may predict or identify diseases with a level of effectiveness comparable to or even exceeding that of physicians. Notably, a comparable positive outlook among the medical community concerning the advantageous uses of AI in routine clinical practice was reflected in studies examining the perspectives of physicians in Portugal, Germany, and other countries of AI. Physicians appear to have a favorable inclination towards the integration of AI in their everyday clinical work, convinced that it will transform their field and elevate the standard of medical care delivered. This advancement is expected to significantly enhance daily clinical tasks, including the management of health information, improving access to and the analysis of health data, as well as reducing medical errors and misdiagnoses [32,33,34,35,36,37,38,39,40,41,42]. Nonetheless, apprehensions were raised regarding job security, future employability, and the privacy of patients [21].

From this survey, essential concerns among healthcare personnel also transpire regarding the implementation of AI in healthcare. Approximately 8% of respondents believed that AI is incapable of providing insights in unforeseen circumstances, while 13.3% voiced the opinion that AI lacks the flexibility required for application to every patient. Additionally, 5.1% indicated that AI struggles to address controversial topics. Importantly, there was a strong consensus among respondents on specific issues, particularly regarding the threat to future employment rates posed by AI. Notably, 57.5% of nursing staff and half of other healthcare professionals expressed concerns about the potential ramifications of AI replacing their jobs and other ethical issues. Interestingly, although nursing staff demonstrate high expectations regarding the potential uses of AI in nursing care (M = 3.99) and its future (M = 4.30), they also recognize the limitations of AI. They perceive that AI may have difficulty offering insights in unforeseen circumstances (M = 1.82), that it lacks the necessary adaptability to be utilized with every patient (M = 1.75), and that it may present challenges when addressing contentious issues (M = 1.58). Furthermore, nursing staff contend that AI possesses a diminished capacity for empathy and for considering a patient’s emotional welfare (M = 1.77). The feedback from other healthcare personnel is somewhat more favorable when compared to that of nursing personnel. Considering the aforementioned, a significant level of skepticism, mistrust, and confusion exists regarding the actual capabilities of AI, leading to the expression of anxiety and the fear of distorting relationships with patients and creating uncontrolled outcomes. The increased occurrence of negative attitudes seems to be associated with a significantly heightened cognitive and behavioral dimension of employee cynicism. This suggests that the characteristics of the workplace may negatively affect healthcare professionals’ views on the implementation of AI in the hospital environment, or that their perspectives are disregarded. Comprehending the apprehensions of healthcare professionals is essential for facilitating a smooth transition to an AI-enhanced healthcare system. Our findings highlight the lack of information, guidance, and structured training of nursing personnel on AI applications in a supportive work environment, which partially explains the limited knowledge of nursing personnel and their reluctance to return some questionnaires distributed in this survey (65/300). In previous studies conducted in Europe, the United States, and other regions, a lack of knowledge, familiarity, and confidence, moral concerns, and ethical issues have been identified and confirmed. Specifically, it has been reported that merely 45% of nurses in Saudi Arabia exhibited a fundamental understanding of AI applications within the healthcare sector. Furthermore, it was disclosed that merely 40% of nurses in the United States expressed confidence in their comprehension of AI algorithms and big data analytics. In China, a study revealed that 57% of individuals had only a limited understanding of AI, while 4.7% were completely unaware of it. Additionally, 64.7% had minimal knowledge regarding AI in the nursing field, and 13.4% were uninformed about AI in nursing. Furthermore, 26.9% and 51.2% of the respondents expressed that the integration of AI into nursing is perceived as very beneficial and beneficial, respectively [17,26,35,43,44,45,46,47,48]. Hence, it was reported that healthcare providers felt that their jobs were at risk of being supplanted by AI in their work environments. The apprehensions and fears surrounding unemployment, combined with the anxiety of needing to acquire new skills, can be profoundly distressing for individuals who have devoted their lives to the healthcare field [49,50,51,52,53]. Beyond these, there exists a distinct necessity for a robust regulatory framework that will establish standards, guidelines, and ethical frameworks pertaining to the application of AI within healthcare environments [40,51,52,53,54,55]. In numerous instances, these obstacles not only impede the integration of AI, but also influence the job satisfaction and commitment of healthcare personnel, thereby making it essential to tackle these issues in order to cultivate a more supportive work environment.

Consequently, it appears that healthcare professionals possess a moderate understanding of AI, and it was identified that there is a clear difference between healthcare professional categories regarding their perceptions of its implications. In terms of healthcare professionals’ familiarity with AI concepts based on sociodemographic characteristics, this study revealed that educational background was a significant factor. It also seems that younger healthcare providers are more familiar with new technologies and AI, which suggests that introducing AI into medical schools would be the best practice for familiarity and readiness. Similarly, it appears that healthcare professionals possessing greater knowledge of AI were more inclined to demonstrate confidence in their positions and a heightened intention to remain. For this reason, investment by institutions in accessible AI tools, along with continuous support, could improve the adoption of and satisfaction with AI. Finally, most participants view the future application of AI optimistically and believe that the advantages surpass the potential risks. This perspective aligns with findings from research conducted in developed nations. A potential explanation for this discrepancy is the prevailing belief among healthcare personnel that AI cannot replicate human emotions or demonstrate empathy, and thus cannot engage in the complex interactions necessary to reassure patients and earn their trust [56,57,58].

6. Conclusions

This survey study evaluated the attitudes and perceptions of healthcare employees regarding the implementation of AI in the healthcare sector. The research was conducted at the largest hospital in Athens, Greece’s capital. The findings were mixed: while the healthcare community is beginning to acknowledge the transformative potential of AI technologies for future outcomes, the integration of AI applications into healthcare is not progressing as swiftly as the technology itself. This discrepancy can be attributed, at least in part, to the reluctance of healthcare personnel to embrace technologies that they do not fully understand, and in some instances, a fear of job displacement due to AI. Additionally, there may be uncertainty about accountability in cases where an AI tool causes an error. Undoubtedly, healthcare providers’ perceptions play a key role in the successful implementation of AI and will influence future societal applications in healthcare. Therefore, there is a pressing need for more thorough and accessible education on AI. Targeted training and organizational support are crucial for promoting the adoption of technology, facilitating the effective implementation of medical technologies. Educational initiatives should be tailored to each specific role to meet the varied needs of healthcare professionals. By providing role-specific training modules, healthcare professionals can enhance their confidence and motivation in utilizing new technologies. It is also essential to tackle the concerns, challenges, and issues related to the implementation of AI in healthcare, as well as the potential of these technologies to enhance healthcare processes and efficiency. Adequate training would increase employees’ understanding of AI and their recognition of its potential benefits in the healthcare sector. Governments and academic institutions can play pivotal roles in promoting the adoption of AI technologies in healthcare.