literature review
-
Bias in Human-AI Collaboration
TL;DR: Human–AI collaboration is now common in both low- and high-stakes settings, but human cognitive biases can distort how people evaluate AI outputs. Biases like automation bias (over-trust), algorithmic loafing (reduced effort), and automation-induced complacency (reduced vigilance) can all weaken oversight and lead to serious errors. Because these flawed decisions can also flow back into…
Brian Grellmann 