What is it about?
When people use AI to make decisions about hiring, lending, criminal justice, or healthcare, they often assume the technology is neutral and objective. This article argues that this assumption is dangerous. Drawing on three established psychological theories, it shows how AI systems can actually make it easier for people to discriminate, not harder. People tend to accept AI outputs that confirm their existing biases and ignore ones that challenge them. They justify discriminatory results by pointing to the algorithm as proof of fairness. And they distance themselves from responsibility when things go wrong, telling themselves they were just following the data. Together these three processes turn AI into a kind of moral cover, allowing people and institutions to perpetuate inequality while genuinely believing they are being objective.
Featured Image
Photo by Nahrizul Kadri on Unsplash
Why is it important?
Most efforts to tackle AI bias focus on improving the technology itself, refining algorithms, improving datasets, adding fairness metrics. This article argues that approach misses the deeper problem. The people using these systems bring psychological tendencies that allow bias to persist even when the tools are technically improved. Without addressing motivated reasoning, system justification, and moral disengagement in users and institutions, even well-designed AI will be used in ways that reproduce inequality. The fix requires both better algorithms and better psychological awareness.
Perspectives
My work has always been concerned with how people justify treating others unequally, and how those justifications can operate below conscious awareness. Writing this piece, I became increasingly struck by how elegantly AI fits into existing psychological pathways for moral self-protection. We already knew that people rationalize bias, defend existing systems, and distance themselves from the harms they cause. What AI adds is a layer of technical authority that makes all three of those tendencies feel more legitimate than ever. What concerns me most is not the bad actor who deliberately uses AI to discriminate. It is the well-meaning institution that genuinely believes it is being fair because it is using an algorithm. That belief is precisely what makes the problem so difficult to see and so hard to address. The solution cannot come from the technology alone. It has to start with understanding ourselves.
Dr Islam Borinca
Read the Original
This page is a summary of: AI as moral cover: How algorithmic bias exploits psychological mechanisms to perpetuate social inequality, Analyses of Social Issues and Public Policy, September 2025, Wiley,
DOI: 10.1111/asap.70031.
You can read the full text:
Contributors
The following have contributed to this page
