Feedback Loop Bias: Why AI Peer Review Systems Amplify Rather Than Correct

AI-assisted peer review systems fail not due to AI limitations, but because their feedback loops lack governance structures to prevent bias amplification. As these systems retrain on skewed user feedback shaped by incentives and trust patterns, biases compound exponentially rather than self-correcting. Organizations deploying AI-assisted peer review must redesign their feedback loops with explicit governance and validation pipelines to prevent systematic bias from becoming embedded in their systems.