This thesis investigates the challenges posed by cognitive biases in the utilization of Large Language Models (LLMs) within the pharmaceutical industry and proposes strategies to mitigate these biases using educational tools aligned with the "Anxious" aspect of the BANI framework. Cognitive biases such as confirmation bias, automation bias, loss aversion, and fear of missing out can significantly impact decision-making, document trustworthiness, and productivity when employees rely on LLMs for generating documents.

To address this issue, this thesis develops and evaluates an E-learning prototype designed to raise awareness of these biases and promote empathy and mindfulness in LLM utilization. Using the Design Science Research (DSR) methodology, the study progresses through the phases: Awareness, where the problem is identified and contextualized within the BANI framework; Suggestion, which identifies E-learning as the most scalable and effective tool to address cognitive biases; Development, where a prototype E-learning course is designed with interactive videos, real-world scenarios, and quizzes; and Feedback & Evaluation, where the prototype is validated through participant feedback.

The findings demonstrate that the E-learning tool successfully increases awareness of cognitive biases and encourages mindful and cautious decision-making, enhancing document trustworthiness while maintaining productivity. While the thesis makes significant contributions, it also acknowledges its limitations. It focuses exclusively on the utilization of LLMs rather than their development, examines only a select few cognitive biases, and tailors its solutions to the pharmaceutical industry. The E-learning prototype, while effective, is a preliminary version that requires further expansion and testing for broader applicability. The research has several implications for management. It highlights the need for organizations to recognize and address cognitive biases as they integrate LLMs technologies into their workflows. By adopting scalable educational strategies like E-learning, companies can foster a culture of critical thinking and mindful engagement with AI tools, improving decision-making and operational efficiency. Future research should expand the scope to include biases in LLMs development, explore other industries, and evaluate the long-term impact of educational interventions on organizational performance. This thesis provides a strong foundation for managing biases in LLM utilization, ensuring ethical, efficient, and reliable use of these technologies in business settings.