Biased Error Attribution in Multi-Agent Human-AI Systems Under Delayed Feedback

arXiv·medium signal

Researchers from Georgia Tech examine how cognitive biases distort human error attribution in multi-agent human-AI systems, particularly under delayed feedback conditions where outcomes are uncertain. The study reveals systematic patterns where humans misattribute errors between AI agents and human operators, with implications for trust calibration, accountability, and safe deployment of multi-agent systems in high-stakes domains.