Exploring Greek Students’ Attitudes Toward Artificial Intelligence: Relationships with AI Ethics, Media, and Digital Literacy

Abstract

This exploratory study (N = 310) investigates the relationship between students’ attitudes toward artificial intelligence (AI), their attitudes toward AI ethics, and their media and digital literacy levels. This study’s specific objectives were to examine students’ (a) general attitudes toward AI, (b) attitudes toward AI ethics, (c) the relationship between the two, and (d) whether attitudes toward AI are associated with media and digital literacy. Participants, drawn from a convenience sample of university students, completed an online survey including four scales: (a) a general attitude toward AI scale (including two subscales, positive and negative attitudes), (b) an attitude toward AI ethics scale (including two subscales, attitudes toward accountable and non-accountable AI use), (c) a media literacy scale, and (d) a digital literacy scale, alongside demographic information. The findings revealed that students held moderate positive attitudes toward AI and strong attitudes favoring accountable AI use. Interestingly, media literacy was positively related to accountable AI use and negatively to positive attitudes toward AI, whereas digital literacy was positively related to positive attitudes, and negatively to negative attitudes toward AI. These findings carry significant theoretical implications by highlighting the unique relationship of distinct literacies (digital and media) with students’ attitudes. They also offer practical insights for educators, technology designers, and administrators, emphasizing the need to address ethical considerations in AI deployment.

1. Introduction

Artificial intelligence (AI) is already shaping human life in diverse ways and is expected to continue significantly impacting areas such as medicine [1], education [2], and retail [3]. AI can be defined as the simulation of human intelligence in machines that can perform certain tasks such as problem-solving or learning, imitating human capabilities [4]. Given AI’s growing influence in education, many studies have explored students’ attitudes toward specific aspects of AI, such as chatbots [5], AI-based assessment systems [6], or ChatGPT [7,8]. This study, however, focuses on students’ general attitudes toward AI rather than specific applications. Understanding general attitudes toward AI is important as the theory of planned behavior [9] suggests they could predict students’ intention to use AI, and consequently, their actual AI use in the future. The primary objective of this study is to explore students’ attitudes toward AI while investigating two important factors that may be related with these attitudes: students’ attitudes toward AI ethics and their levels of media and digital literacy.

The first specific objective of this study is to explore general attitudes toward AI among a university student sample. While AI is often praised for its transformative potential, it also raises ethical concerns [10]. Students may experience unease over issues like transparency (is AI functioning open and clear to users?), fairness (is AI free of bias?), and accountability (who bears responsibility for its use?). The second specific objective is to explore students’ attitudes toward AI ethics, focusing on their perceptions of these critical ethical dimensions. The third specific objective is to analyze the relationship between general attitudes toward AI and attitudes toward AI ethics. We expect a positive correlation between the two so greater ethical concerns will be correlated with less positive general attitudes toward AI.

This study also investigates the role of students’ literacy skills, specifically media and digital literacy [11], as a key predictor of their general attitudes toward AI. Understanding this relationship is important because students’ perceived ability to critically engage with media (media literacy) and navigate the digital landscape (digital literacy) may affect their overall evaluation of AI. The fourth specific objective of this study is to explore the link between students’ general attitudes toward AI and their literacy skills, thus providing a deeper understanding of the factors shaping students’ perspectives on AI.

1.1. Attitudes Toward AI

Public acceptance of generalized AI use and its integration into daily life largely depends on individuals’ general attitudes toward AI, that is, their evaluations of AI’s potential benefits or drawbacks [4,12,13]. Research in the social sciences, particularly psychology, has shown that attitudes are powerful predictors of behavior [9], making the study of general attitudes toward AI essential for understanding students’ views and developing effective AI-related policies. Prior research reveals mixed attitudes toward AI: positive attitudes often derive from AI’s potential to help in certain tasks and decision-making, while negative attitudes tend to focus on perceived threats, such as impacts on job security [14,15]. Many studies, including those by Schepman and Rodway [13,16], indicate that general attitudes are somewhat positive but moderate, with responses clustering around the midpoint on attitudinal scales (e.g., positive attitudes had an average mean of 3.5, while negative an average mean of 3.1—on a 5-point scale, 6, 10). A similar pattern of results was also evident in a study by Sindermann et al. [17], particularly in the European samples examined.

Several studies across diverse countries and cultural contexts demonstrate the growing empirical interest in students’ attitudes toward various aspects of AI, indicating generally positive views toward AI, often accompanied with specific concerns. For example, Acosta-Enriquez et al. [7] found that among Peruvian college students, the emotional dimension of attitudes was more significant than the cognitive or conative ones. Similar patterns were also observed in Greek social science students [18] and in Spanish students across various fields who expressed positive attitudes and a desire for more AI-focused education [19]. In Taiwan, students appreciated AI’s role as a tool and tutor in learning contexts [20], while in India, management students held positive attitudes and favored increased AI integration in education [8].

Conversely, some studies reveal more complex views among students, often depending on their academic field, demographic characteristics, and literacy levels. For example, Filipino students’ attitudes varied by academic field and perceived literacy [21]. In Slovenia, social science students and male students showed more positive attitudes than computer science students and female students [22], a pattern also observed in a Swedish sample [5]. Additionally, two multinational studies provide further insights into students’ mixed attitudes: one study conducted in Australia, Cyprus, and the U.S. found mixed attitudes toward AI in assessment, largely due to concerns about AI’s potential negative impact on creativity [6]. Another study across Iraq, Kuwait, Egypt, Lebanon, and Jordan found that positive attitudes were influenced by perceived ease of use and usefulness [23]. Overall, these studies suggest that, while students tend to have positive attitudes toward AI, these attitudes are often qualified by various contextual factors and concerns [24,25]. Because attitudes toward AI are important in predicting AI acceptance and use, the present study aims to further our understanding of students’ attitudes toward AI by contributing to the growing body of literature on the subject by focusing on a Greek university context.

1.2. Attitudes Toward AI Ethics

Beyond general attitudes, ethical considerations also significantly influence how students view AI. Common concerns include potential biases in AI decision-making (e.g., discrimination based on gender or race) and issues of accountability and transparency—such as whether AI’s decision-making processes are understandable and clear to users [26]. The ethics of AI has attracted attention from a variety of public institutions (from enterprises to NGOs), which have identified key issues such as transparency, fairness, privacy, responsibility, and non-maleficence [10]: Transparency (or explainability) addresses an AI’s ability to render its decision-making processes comprehensible to human users [27], while fairness questions whether the AI displays biases that could lead to prejudice and social exclusion [28]. Privacy pertains to secure data handling, responsibility concerns accountability of AI functioning, and non-maleficence focuses on AI’s potential to cause harm [29]. Given the importance of users’ attitudes toward AI ethics, this study also examines whether students’ ethical considerations relate to their general attitudes toward AI.

1.3. Media and Digital Literacy

The current study also investigates whether students’ perceived ability to navigate the digital world and critically evaluate media content relates to their attitudes toward AI. Literacy, in this context, encompasses the knowledge, competencies, and attitudes gained through media and ICT education [30]. Although literacy has been conceptualized in multiple ways [31], this study focuses on two central types: media and digital literacy. Media literacy involves critical engagement with media content and analysis of media’s societal role [11], while digital literacy refers to competencies in using digital technologies [32]. Media literacy is grounded on a critical view of the role of media in society, emphasizing its economic and ideological functioning. Media language is assumed to affect the construction of meaning, which is nevertheless negotiated by receivers [33]. In other words, media literacy is associated with individuals’ abilities to critically consume, question, and analyze information. Digital literacy is associated with people’s ability to constantly adapt to new technologies [34] and focuses on competencies related to the digital and internet world. This focus on technology, however, has attracted criticism for overemphasizing functionality and information retrieval [35]. This study explores whether media and digital literacy independently predict students’ attitudes toward AI. This is a question that has not been empirically tested yet.

1.4. Research Questions and Hypothesis

Based on the discussion above, this study investigates the following research questions (RQ):

• RQ1a: what are students’ general attitudes toward AI;
• RQ1b: what are students’ attitudes toward AI ethics.

To examine general attitudes toward AI, this study employs Schepman and Rodway’s [13,16] scale of general attitudes, which measures emotional reactions, evaluations of societal and personal utility, and concerns. This scale has two subscales of positive and negative attitudes. The former captures positive affective evaluations toward AI (example item, AI is exciting), attitudes regarding intentions to use AI in everyday life (example item, I am interested in using artificially intelligent systems in my daily life), or attitudes toward AI performance (example item, artificially intelligent systems can perform better than humans). The negative subscale captures negative affective evaluations (example item, I think artificial intelligence is dangerous), attitudes toward performance issues (example item, I think artificially intelligent systems make many errors), etc. The scale has shown satisfying predictive and convergent validity against relevant measures. This is the first study to apply Schepman and Rodway’s scale to a Greek university sample (see [18] for a study employing a different scale), something that could enhance our understanding of this specific student population.

To examine attitudes toward AI ethics, this study employs Jang et al.’s [26] scale that captures concerns regarding five issues: transparency (is AI functioning open to users?), fairness (is AI free of bias?), privacy (is AI’s handling of data safe?), responsibility (who is accountable for the AI’s functioning?) and non-maleficence (can AI harm humans?). The scale has demonstrated strong psychometric qualities.

This study also proposes the following hypothesis regarding the relationship between general attitudes toward AI and attitudes toward AI ethics.

• H1: general attitudes toward AI will positively correlate with attitudes toward AI ethics.

The relationship between these attitudinal concepts has yet to be studied. We assume that the more ethical (in different dimensions) students view AI use, the more positive attitudes they will hold toward it, and vice versa.

Additionally, this study explores the following research question regarding the relationship between media and digital literacies and general attitudes toward AI.

• RQ2: do media and digital literacy relate to general attitudes toward AI

The relationship between general attitudes toward AI and media and digital literacy has not attracted empirical attention yet. To assess media literacy, this study applies Inan and Temur’s [36] scale, which captures the role of media in constructing reality and having economic, political, and ideological implications as well as the audiences’ ability to negotiate and co-construct the meaning of media content. To examine digital literacy, we used Hargittai and Hsieh’s [37] scale that captures users’ understanding of internet-related terminology and navigational abilities. This study aims to clarify whether these literacies predict general attitudes toward AI and contribute unique explanatory power over and beyond attitudes toward AI ethics.

2. Materials and Methods

2.1. Sample of This Study

A convenience sample of 311 Greek students at a large public university (226 female, 83 males, ages ranging from 18 to 34 years, M_age = 23.0, SD = 6.34) participated voluntarily in an online survey during March and April of 2024. Convenience sampling allowed researchers to promptly gather data from readily accessible participants of the population and could provide, through an initial exploration of the study’s research questions and hypothesis, useful insights to guide future research with more representative samples. However, because of this sampling method’s non-probabilistic nature, generalizations should be made with great caution.

2.2. Research Design

This study employed a correlational, cross-sectional design to address the research questions and hypotheses. This design, which allows for the simultaneous collection of multiple variables, is particularly useful for describing key characteristics of the student population being studied and for exploring potential relationships among the variables of interest.

2.3. Measures

The questionnaire measured three concepts, exploring general attitudes toward AI, attitudes toward AI ethics, and literacy levels. All variables were measured with Likert scales to capture participants’ nuanced responses by providing a range of response options. Likert scales offer a standardized and reliable way to measure and quantify the concepts of interest.

2.3.1. Attitudes Toward Artificial Intelligence

The General Attitudes toward Artificial Intelligence Scale (GAAIS, [13]) was employed. Participants were asked to indicate how they evaluate AI on a scale of 20 items (5-point scale ranging from 1 = strongly disagree to 5 = strongly agree). To reduce the large set of items into a smaller number of components, identifying underlying patterns, a principal component analysis was used. It revealed two factors, the first, positive attitudes toward AI, consisted of 12 items, showing positive general attitudes toward AI (e.g., artificial intelligence is exciting, Cronbach’s α = 0.884, eigenvalue = 6.50, explaining 35% of the variance, all loadings > 0.5). The second, negative attitudes towards AI, consisted of 7 items showing negative general attitudes toward AI (e.g., I find artificial intelligence sinister, Cronbach’s α = 0.834, eigenvalue = 2.75, explaining 15% of the variance, all loadings > 0.5). One item was left out because of a very low loading.

2.3.2. Attitudes Toward Artificial Intelligence Ethics

Participants completed the Attitudes toward Artificial Intelligence Ethics Scale (AT-EAI, [26]), indicating their attitudes toward AI ethics. The scale consisted of 17 items (5-point scale ranging from 1 = strongly disagree to 5 = strongly agree). A principal component analysis revealed two factors, the first, accountable AI, consisted of 9 items indicating that AI logic and functioning should be transparent and accountable (e.g., it is essential for AI to explain the reasons for its decisions, Cronbach’s α = 0.749, eigenvalue = 3.47, explaining 25% of the variance, all loadings > 0.5). The second, non-accountable AI, consisted of 5 items indicating that transparency and accountability are not important for AI functioning (e.g., even if the AI does not explain why it made a particular decision, it is preferable to have higher accuracy, Cronbach’s α = 0.543, eigenvalue = 1.59, explaining 14% of the variance, all loadings > 0.5). Three items were left out of the scales because of very low loadings.

2.3.3. Media Literacy

Participants completed a four-item scale of media literacy (5-point scale, 1 = strongly disagree to 5 = strongly agree, sample item: I would caution people around me about the negative sides and negative effects of media). The scale was based on a scale by Inan and Temur [36], maintaining the core logic of the scale and its dimensions (see also [38]). Principal component analysis revealed one factor (Cronbach’s α = 0.588, eigenvalue = 1.74, explaining 45% of the variance, all loadings > 0.5).

2.3.4. Digital Literacy

Participants completed a ten-item scale of digital literacy [37], indicating their familiarity with internet-related terms such as PDF, phishing, and tagging (5-point scale, 1 = strongly disagree to 5 = strongly agree). Principal component analysis revealed one factor (Cronbach’s α = 0.891, eigenvalue = 5.10, explaining 51% of the variance, all loadings > 0.6).

2.4. Methods of Analyses

To address the research questions and hypothesis, we used descriptive statistics to analyze RQ1a and RQ1b, while correlational and hierarchical regression analyses were conducted for H1 and RQ2.

3. Results

3.1. Preliminary Analyses

To answer RQ1a and RQ1b, descriptive statistics were employed showing that participants held moderately more positive (M = 3.15, SD = 0.71) than negative (M = 2.84, SD = 0.81) attitudes toward AI (RQ1a, see

Table 1. Means, standard deviations, and intercorrelations among variables.

Variables	M	SD	1	2	3	4	5
1. Positive attitudes	3.15	0.71
2. Negative attitudes	2.84	0.81	−0.37 ***
3. Accountable AI use	4.43	0.49	−0.18 **	0.25 ***
4. Non-accountable AI use	2.42	0.58	0.09	−0.06	−0.12 *
5. Media literacy	3.09	0.67	0.06	0.08	0.16 **	0.01
6. Digital literacy	2.98	0.95	0.42 ***	−0.19 ***	-0.09	0.03	0.30 ***

* p < 0.05, ** p < 0.01, *** p < 0.001.

for means, standard deviations, and intercorrelations among variables). This difference was significant [F(1, 310) = 19. 1, p < 0.001, η_p² = 0.06]. Participants held explicitly more positive attitudes toward accountable AI use (M = 4.43, SD = 0.48) than attitudes toward non-accountable AI use [M = 2.42, SD = 0.56, F(1, 310) = 895.0, p < 0.001, η_p² = 0.86, RQ1b]. To examine H1, a correlational analysis was employed, showing, as expected, a positive correlation between attitudes toward accountable AI use and negative attitudes (r = 0.25, p < 0.001) and a negative correlation between attitudes toward accountable AI use and positive attitudes (r = −0.18, p < 0.01). Moreover, participants considered themselves as moderately literate both digitally (M = 3.09, SD = 0.67) and in terms of media (M = 2.98, SD = 0.95). An inspection of correlations among the variables shows that positive attitudes toward AI were positively related to digital literacy and negatively with attitudes toward accountable AI (RQ2). Reversely, negative attitudes toward AI were negatively related to digital literacy and positively related to attitudes toward accountable AI. Interestingly, regarding attitudes toward AI ethics, only attitudes toward accountable AI were related to media literacy. The age of the participants was positively related to media literacy and negatively to negative attitudes.

3.2. Hierarchical Regressions

To answer RQ2 and further examine the predictive power of the variables on attitudes toward AI, two hierarchical regression analyses were performed, where the predictors were successively entered in the analysis (see

Table 2. Summary of hierarchical regression analysis for variables predicting positive attitudes toward AI (N = 310).

	β	95% CI for β		B	p
		LL	UL
Step 1
Gender	−0.71	−0.95	−0.47	−0.50	<0.001
Age	0.03	−0.08	0.13	0.00	0.611
Step 2
Gender	−0.64	−0.89	−0.40	−0.46	<0.001
Age	0.03	−0.07	0.14	0.00	0.527
Accountable AI	−0.11	−0.22	−0.01	−0.16	0.040
Non-accountable AI	0.06	−0.05	0.16	0.07	0.297
Step 3
Gender	−0.36	−0.60	−0.11	−0.25	0.005
Age	−0.03	−0.13	0.07	0.01	0.527
Accountable AI	−0.10	−0.20	0.01	−0.14	0.073
Non-accountable AI	0.07	−0.02	0.18	0.09	0.133
Media literacy	−0.04	−0.15	0.06	−0.05	0.417
Digital literacy	0.37	0.26	0.49	0.28	<0.001

Note. R² = 0.10 (p < 0.001) for step 1; R² = 0.12 (p < 0.001) ΔR² = 0.02 for step 2 (p < 0.05); R² = 0.23 (p < 0.001) ΔR² = 0.11 for step 3 (p < 0.001).

and

Table 3. Summary of hierarchical regression analysis for variables predicting negative attitudes toward AI (N = 310).

	β	95% CI for β		B	p
		LL	UL
Step 1
Gender	0.16	−0.09	0.41	0.13	0.203
Age	−0.11	−0.22	−0.01	−0.01	0.043
Step 2
Gender	0.05	−0.12	0.30	0.04	0.689
Age	−0.12	−0.23	−0.01	−0.01	0.029
Accountable AI	0.24	0.13	−0.35	0.40	<0.001
Non-accountable AI	−0.03	−0.14	0.07	−0.05	0.540
Step 3
Gender	−0.09	−0.35	0.18	−0.06	0.526
Age	−0.09	−0.21	0.01	−0.01	0.085
Accountable AI	0.21	0.10	0.32	0.36	<0.001
Non-accountable AI	−0.05	−0.15	0.06	−0.06	0.399
Media literacy	0.11	0.11	0.23	0.14	0.049
Digital literacy	−0.20	−0.32	−0.07	−0.17	<0.001

Note. R² = 0.02 (p < 0.05) for step 1; R² = 0.08 (p < 0.001) ΔR² = 0.09 for step 2 (p < 0.001); R² = 0.11 (p < 0.001) ΔR² = 0.04 for step 3 (p < 0.01).

). In the first step of the regression analyses, control variables were entered, that is, gender and age. In the second step, the attitudes toward AI ethics (accountable and non-accountable AI) were entered. In the third step, the two literacy variables (media and digital) were entered. All steps were statistically significant.

The first analysis showed that digital literacy positively predicted positive attitudes toward AI. Attitudes toward accountable AI use were negatively related, and men had more positive general attitudes than women. The second analysis showed that attitudes toward accountable AI use and media literacy, positively, and digital literacy, negatively, predicted negative attitudes toward AI.

4. Discussion

This study’s primary objective was to examine university students’ general attitudes toward artificial intelligence (AI) and the relationship with their attitudes toward AI ethics and media and digital literacy. It explored how students generally evaluate AI and whether such evaluation is related to the ethical concerns they may have about AI use and to their ability to critically engage with media content and navigate the digital world. Regarding the first specific objective of this study, the exploration of students’ general attitudes toward AI, key findings reveal that young individuals hold more positive (than negative) attitudes toward AI (RQ1a). A closer look at the actual results showed that both positive and negative attitudes were relatively close to the midpoint of the measurement scale, indicating that, although participants generally showed positive attitudes, both positive and negative attitudes did not deviate significantly from a moderate stance. These results align with previous studies [13,16], showing that university students hold a moderately positive view of AI, demonstrating a tendency to accept AI and recognize its potential benefits.

Interestingly, although positive attitudes were not related to students’ age, negative attitudes were negatively related to age, indicating that younger students tended to have more negative attitudes. This is compatible with other studies showing that older students were using AI more often than young adults [26,39]. These younger students also perceived themselves as less digitally literate [see also 26] as age and digital literacy—but not media literacy—were positively related (r = 0.21, p < 0.001). Moreover, the regression analysis showed that men tended to have more positive attitudes toward AI than women, a finding that may be related to men’s greater perceived digital literacy [M_men = 3.56, SD = 1.03, vs. M_women = 2.77, SD = 0.82, t(1, 309) = 7.02, p < 0.001]. For both age and gender, digital literacy seems then to be an important factor in predicting students’ general attitudes. Overall, it is evident that university students tend to have positive attitudes, which are qualified by their demographic characteristics and perceived literacy, a finding that provides a more nuanced view of this topic.

This study’s second specific objective was to examine students’ attitudes toward AI ethics. Participants clearly favored accountable (vs. non-accountable) AI use (RQ1b). They favored AI’s decision-making to be transparent and open to its users, as is commonly found in similar studies, e.g., [27,40], and do not want to outsource the responsibility for AI use to technological developers. This is a clear normative concern that should guide future AI technology development. While gender was not related to the attitudes on non-accountability, there was a significant difference in accountable attitudes, with women being more favorable to transparent AI use than men [M_men = 4.28, SD = 0.45, vs. M_women = 4.49, SD = 0.48, t(1, 309) = −3.44, p < 0.001]. These results reinforce prior findings [2] that found that women can be more sensitive in various dimensions of AI ethics. No significant relationship was found between attitudes toward AI ethics and age.

Moreover, this study’s third specific objective was to examine the relationship between attitudes toward AI ethics and general attitudes toward AI (H1). As expected, accountable attitudes toward AI ethics were positively related to negative attitudes and negatively to negative attitudes. The more positive the attitudes the participants had toward AI, the less accountable they thought that AI use should be. The reverse pattern was observed for negative attitudes. This pattern of results was evident not only in the correlational analysis but also in the hierarchical regression analysis, where other factors (demographics and literacies) were also included and controlled for, providing further support for this relationship. These findings also emphasize the importance of addressing students’ attitudes toward AI ethics when developing AI or AI-related educational programs.

This study’s fourth objective was to investigate the role of media and digital literacy in predicting students’ attitudes toward AI (RQ2). The regression analyses showed that digital literacy was positively related to positive attitudes and negatively to negative attitudes toward AI. The more confident participants felt in navigating the digital world, the more positively—and less negatively—they viewed AI. Interestingly, when all factors were included in the regression analysis, the perceived ability to critically consume media content predicted general negative attitudes toward AI; the more participants believed they had the competencies to critically consume media content, the more negatively they viewed AI. Media literacy was also positively related to accountable attitudes toward AI ethics. These findings suggest, for the first time in the literature, that people’s perceptions about their own capabilities to use media are important predictors of their attitudes. Media and digital literacy are differentially related to attitudes toward AI, highlighting literacy as an important factor in predicting general attitudes.

4.1. Theoretical and Practical Implications

This study enhances our understanding of students’ general attitudes toward AI and AI ethics by using psychometrically robust measures in a cultural context that has not been extensively studied in the past. It also sheds light on the relationship between users’ abilities and their attitudes toward AI. While previous studies have focused on students’ perceived efficacy, e.g., [19,22], as a predictor of their attitudes, this study also examines perceived media and digital literacy as relevant but distinct predictors. The differing associations of literacies with attitudes (with digital literacy positively related to positive attitudes and negatively to negative attitudes, while media literacy is positively related to negative attitudes) offer a more nuanced understanding of these relationships. This study provides practical implications, especially given students’ strong concerns about the ethical use of AI. Educators should consider these concerns when designing pedagogical approaches, emphasizing AI’s benefits while addressing potential risks. AI-based education must focus on improving AI literacy within the student community, equipping them with critical skills for safe and productive use. Furthermore, technology designers should prioritize creating safe, user-friendly interfaces that address users’ ethical concerns. Lastly, administrators could develop policies that promote and support ethical AI use.

4.2. Limitations

The present study has certain limitations. First, it employed convenience sampling, which undermines the generalizability of the findings. A more representative sample would more confidently support these findings. Second, the overrepresentation of female students in the sample may have influenced the findings since women were found to hold less positive attitudes toward AI. Third, while a culturally diverse sample was employed in this study, no culturally relevant measures were taken to identify the potential importance of such factors in affecting the results. Fourth, other factors that could have affected these results, such as prior AI experience or education, were left out of this study even though they have been found to play a role in the prediction of attitudes toward AI, e.g., [41]. Despite these limitations, the findings of the present study contribute to the growing literature on people’s views of AI. These attitudes were shown to be related to their attitudes toward the ethical aspects of AI as well as to participants’ perceived abilities in consuming media content and navigating the digital world.

5. Conclusions

This study provides insights into university students’ general attitudes toward AI as well as their attitudes toward AI ethics. The findings reveal predominantly positive but moderate general attitudes. Ethical concerns, particularly around accountability, are prominent, with women showing a stronger preference for transparency in AI. Digital and media literacy emerged as important predictors, with digital literacy associated with positive attitudes and media literacy with negative ones. These findings underscore the need for targeted technology and educational initiatives that address both AI’s potential and ethical complexities.