Investigating Pharmacists Perceptions and Engagement with Generative AI in Everyday Contexts and Decision-Making

INTRODUCTION

Generative Artificial Intelligence (Gen AI) refers to the system that capable of creating text images or other media forms through utilization of various generative models (like deep learning and large Language Model techniques). Models acquire patterns and structures within training data subsequently generating new data through complex algorithmic manipulations very quietly. Figure 1 represents Gen AI encompasses myriad (Multi) task types and forms of media generation rapidly across diverse specific areas. Generative AI possesses diverse content creation capabilities and human-like conversational prowess alongside data augmentation abilities typically. Engagement entails active involvement of individuals in daily life while making somewhat informed decisions regarding overall well-being quite effectively. Individuals ‘Pharmacists perceptions are reflected quite vividly in their level of engagement. Basic outcomes of an engagement strategy include value and efficiency and trust sometimes but also consistency relevance and control simultaneously. Generative AI offers diverse content creation capabilities and enables human-like conversations while generating vast amounts of training data through complex data augmentation techniques. Goldman Smith asserts global economy grew at 7% rate over past decade and that feat starkly illustrates Gen AI's profound impact on people's daily perceptions. Gen AI has myriad uses across various sectors nowadays quite effectively.

Click to enlarge

Review of Literature

Gen AI models, such as ChatGPT and Gemini, provide powerful generative capabilities, they may still lack the reliability required for high stakes in decision-making purposes particularly in the individual perception. Meanwhile, the rapid adoption of Gen AI has encouraged a re-examination of existing ethical, legal and regulatory frameworks, as its ability to generate original content challenges the traditional system of authorship, ownership and accountability in the decision-making processes. The exploring of the public perception towards Generative AI is crucial and to acceptance and trust is the significant shape, acceptance and development of applications of AI technologies across the essential area of the study (Sharma, 2024). According to the Public Engagement with Generative AI depends on the degree of the individuals feel empowered and secure while using AI tools (Dwivedi et al., 2023). And Trust plays a significant role in the technology acceptance, especially the technology performs complex cognitive tasks, such as content generation and decision making (Venkatesh et al., 2003). However, for Gen AI, trust, transparency, and explainability of additional critical factors (Shin, 2021). Most of the researchers said that Passive AI usage on active engagement, especially when the users perceive the AI outputs to be accurate their needs and preferences (Siau and Wing, 2018). The black box and the nature of Generative AI models concerns decision making and transparency and algorithmic bias (Binns, 2018). In educational and Professional settings, Generative AI is increasingly viewed as productivity enhancer, especially when it contributes to the cognitive tasks, and problem-solving (Zhai, 2022). However, there is a still a gap in public understanding of how these systems function, which can result in misuse (Brennen et al., 2023). The emotional and ethical Pharmacists perceptions of users significantly influence their willingness to use Gen AI tools for decision-making. The users believe that the AI respects their privacy, maintain data security, and provides equitable results, they are more likely to adopt it (Zhou et al., 2021). The relationship between perception, trust, and engagement is interdependent. While users may to find Gen AI tools engaging and useful, decision-making reliance is still evolving, depending on Pharmacists personal experiences, technological literacy, and cultural context (Lu et al., 2021).

Conceptual Model

Conceptual Model illustrates a conceptual model entailing relationships between independent variables namely Perception of Generative AI and Trust in Generative AI alongside Decision making influence and their direct bearing on dependent variable Engagement with Generative AI. Individuals Pharmacists perceptions of generative AI and its impact on decision-making processes collectively influence their level of engagement with such tools quite significantly. This framework lays a foundation beneath understanding key psychological drivers and functional motivations sparking user interaction with Generative AI tools heavily nowadays.

Click to enlarge

MATERIALS AND METHODS

A total of 180 responses were generated from a google form and analysis was done through JAMOVI software desktop version used entirely for this study. Reliability test yielded over 81.5% and must be accepted for study while exploratory factor analysis validated each variable thoroughly.

RESULTS

Data Analysis

Table 1 represents the reliability statistics for the scale measuring the constructs used in the study. The Cronbach’s α value of 0.815 and McDonald’s ω of 0.822 indicate a strong level of internal consistency across the items. These values suggest that the items reliably measure the underlying construct with minimal measurement error. Overall, the scale demonstrates satisfactory reliability, confirming its suitability for further statistical analysis. Table 2 summarizes the exploratory factor analysis results for the Public Perception construct. All factor loadings range between 0.621 and 0.719, demonstrating strong item-to-factor relationships. The Bartlett’s Test of Sphericity is significant (p<0.001), and the overall KMO value of 0.731 indicates good sampling adequacy. These results confirm that the items meaningfully represent a single underlying factor. Thus, the Public Perception scale is both valid and statistically robust. Table 3 presents the factor analysis results for the Trust in Generative AI construct. The factor loadings, ranging from 0.674 to 0.800, show that each item contributes strongly to the underlying factor. A significant Bartlett’s Test (p<0.001) and a KMO value of 0.744 indicate that the dataset is suitable for factor analysis. These findings demonstrate that the trust construct is well-defined and structurally sound. Overall, the scale exhibits strong validity and item cohesion. Table 4 displays the factor loadings for Engagement in Generative AI, with values ranging from 0.569 to 0.725. The Bartlett’s Test of Sphericity is significant (p<0.001), and the overall KMO value of 0.742 shows adequate sampling quality. Although two items have moderate loadings, the construct still retains acceptable validity. These results suggest that engagement-related behaviors cluster reasonably well under a single factor. Thus, the scale is statistically appropriate for further analysis. Table highlights the factor analysis results for Decision Making using Generative AI. The factor loadings range from 0.606 to 0.699, indicating acceptable but comparatively weaker item associations. The Bartlett’s Test is significant (p<0.001), confirming factorability, while the KMO value of 0.649 falls slightly below the preferred threshold of 0.70. This suggests marginal sampling adequacy and a weaker factor structure. Therefore, the validity of this construct may be limited and may require improvement or item refinement.

Table 5 represents the Multiple regression model summarized indicates an R value 0.630 and R² value of 0.397, meaning that 39.7% of the variation in engagement is explained by Perception, Trust, and Decision Making. Table results show that Perception (β=0.312, p<0.001) and Decision Making (β=0.204, p=0.042) significantly predict engagement, while Trust does not reach significance. The model demonstrates that perception and decision-related factors contribute meaningfully to engagement with generative AI tools. The regression equation reflects these relationships clearly.

DISCUSSION

The study demonstrated strong reliability, with a Cronbach’s alpha of 0.815 confirming acceptable internal consistency of the scale. Exploratory Factor Analysis supported data validity, although one decision-making item showed a weak loading and may require revision. Regression results revealed a moderately strong overall relationship (R=0.631), suggesting meaningful associations among the key variables. Findings indicate that generative AI tools have a low but noticeable impact on trust, reflecting pharmacists’ cautious attitudes toward AI reliability. However, AI showed a stronger influence on decision-making, helping pharmacists improve judgment and efficiency. Subtle contextual factors continue to shape perceptions, especially concerning trust. Overall, generative AI tools significantly affect pharmacists’ perceptions and contribute to enhanced decision-making practices.

FINDINGS

Study findings involve a reliability test yielding Cronbach's alpha value of 0.815 indicating acceptable internal consistency level and Exploratory Factor analysis was subsequently conducted for assessing data validity. All items loaded strongly into respective factors except one item related to decision making with AI showed a woefully weak loading. Regression analysis was conducted fairly thoroughly between several key variables under various conditions. Overall R-value stood at 0.631 indicating moderately strong correlation deemed fairly acceptable in most statistical analyses nowadays. A regression equation was formulated rather haphazardly amidst efforts undertaken subsequently to gain better insight into various underlying factors. Results indicate low impact mainly on trust and somewhat on decision-making dimensions heavily influenced by various subtle factors. Generative AI tools significantly influence individual Pharmacists perceptions thereby bringing about lifestyle changes and bolstering decision-making faculties remarkably well.

CONCLUSION

Gen AI tools are being increasingly utilized daily largely due to public perception playing a major role in shaping various engagement levels effectively. Gen AI creates diverse content remarkably well and simplifies tasks somewhat but raises concerns around trust and reliability in decision-making processes. Regression value suggests a relationship exists between perception and engagement, but trust factors and decision-making processes need bolstering greatly now. Full confidence in outputs from Gen AI for crucial decisions remains somewhat elusive nowadays. Users get engaged pretty quickly and find tools super innovative, but lack of trust and transparency hinders deeper adoption significantly. Fully integrating generative AI into daily existence necessitates markedly enhancing explainability and trust essentially across the board suddenly nowadays.

SUGGESTIONS AND FUTURE SCOPE

Developers should prioritize transparency and ensure outputs are ethically designed and fairly explainable somehow for public trust in Generative AI. Awareness initiatives must be launched swiftly nationwide informing users about Gen AI tools' efficacy and inherent limitations pretty effectively. Policymakers and developers must intricately collaborate establishing nebulous guidelines on murky ethical use and stringent accountability measures pretty quickly. Incorporating feedback mechanisms and crafting intuitive interfaces can rather significantly boost user confidence quite remarkably and enhance overall engagement gradually. Institutions organize workshops quite frequently for training individuals on responsible Gen AI usage in personal contexts and various other professional settings effectively. Efforts must be made carefully localizing AI tools suitable for diverse cultural contexts and varying linguistic preferences enhancing adoption rates significantly.