Explainable Artificial Intelligence (XAI) from a user perspective- A synthesis of prior literature and problematizing avenues for future research

Explanations need four key components to be effective: AI explanations must have the right format (how it's presented), completeness (all necessary information included), accuracy (correct information), and currency (up-to-date details). Users should also be able to ask for additional information beyond what's automatically provided.

XAI impacts five critical user outcomes: Explainable AI affects trust in the system, transparency about how it works, whether users understand it, how usable it is, and perceptions of fairness—all essential for responsible AI deployment.

One-size-fits-all explanations don't work: The research emphasizes that explanation format and content should be adaptable to user needs, suggesting that effective XAI requires considering who the user is and what they need to know, rather than providing static explanations.

There's a gap between XAI research and real user needs: This systematic review of 58 studies reveals that while XAI research exists, there's limited understanding of how explanations actually affect user behavior and decision-making in practice—a critical gap for ensuring AI systems are truly responsible.

Fairness is inseparable from explainability: The framework identifies fairness as a direct effect of XAI, indicating that making AI systems more explainable is not just about transparency but is fundamental to addressing bias and ensuring equitable outcomes.