Artificial Intelligence and Cognitive Diversity: Exploring Implications for HR and Organizational Management

Amid the rapid advancement of artificial intelligence (AI), organizations are continuously confronted with new challenges. This paper delves into the impact of discriminative and generative AI on organizational and societal diversity. To replicate and extend previous research (e.g., Atari et al., 2023), which found that ChatGPT and other generative AI models exhibit WEIRD cultural biases, across various generative AI models, including GPT-4o, Llama 3.2, EEVE, and EXAONE 3.5. The findings reveal discrepancies in algorithmic transparency and the extent of Western-centric data training across these models. While some models show a reduction in bias compared to earlier versions, notable gaps remain in representing certain cultural contexts. AI is categorized into discriminative and generative types.. Discriminative AI enhances predictive and classification capabilities through supervised learning, facilitating automation and personalized services; however, the accumulation of data bias poses a threat to organizational diversity. Generative AI, by leveraging unsupervised and reinforcement learning, generates creative ideas and content, fostering innovation but simultaneously carrying the risk of value-laden biases stemming from training data. This study provides a detailed analysis of the potential impacts and risks of both AI types across HR processes, including recruitment, promotion, and performance evaluation. Furthermore, it incorporates regulatory and ethical frameworks related to human capital disclosure and AI governance, addressing key considerations for organizations in ensuring AI transparency and fairness. By highlighting the complementary use of discriminative and generative AI, this paper suggests pathways for organizations to achieve creativity, operational efficiency, and inclusiveness, contributing valuable theoretical and practical insights.