Critical Race Theory emerged in American legal academia in the late 1970s, developed by scholars who were frustrated that the formal legal victories of the civil rights movement had not produced the substantive racial equality they promised. Laws prohibiting discrimination existed. Discrimination persisted. Something was wrong with the framework being used to understand racism.
The founding insight of CRT is simple but radical: racism is not primarily a matter of individual prejudice. It is structural. It is embedded in institutions, systems, laws, and practices in ways that produce racially unequal outcomes even when no individual actor holds explicitly racist views. A hiring process can discriminate without any discriminatory hiring manager. A lending algorithm can redline without any banker consciously thinking about race. A school funding system can produce racially unequal educational outcomes without any policymaker intending that result.
The scholars who developed CRT, Derrick Bell, Kimberlé Crenshaw, Mari Matsuda, Richard Delgado, Patricia Williams among others, drew on both legal scholarship and the lived experience of people of colour to develop a set of core analytical commitments:
Racism is ordinary, not aberrational. Racist outcomes are the normal state of affairs in societies built on racial hierarchy, not exceptional deviations from an otherwise fair system. This means that the question is not “was there racism here?” but “how is racism operating here?”
Interest convergence. Derrick Bell observed that civil rights advances for Black Americans tended to occur when they converged with the interests of white Americans. Progress on racial justice was rarely won on its merits alone. This insight applies directly to AI ethics reform, when does it happen, and whose interests does it serve?
The social construction of race. Race is not a biological category. It is a social and legal construction that has real material effects. AI systems that treat racial categories as natural and stable are building on a false foundation.
Intersectionality. Kimberlé Crenshaw’s foundational contribution: race does not operate alone. The experience of a Black woman is not simply the experience of being Black plus the experience of being a woman. It is a distinct experience shaped by the intersection of race and gender, and other axes of identity. AI systems designed to address one dimension of inequality while ignoring others will systematically fail the people at the intersections.
Counter-storytelling. The experiences of people of colour are systematically excluded from dominant narratives about how systems work and who they serve. CRT foregrounds these excluded voices as a methodological and political commitment.
Why does this belong in an AI ethics curriculum? Because AI systems are not built in a vacuum. They are built by people, trained on data, and deployed in societies that carry the full weight of racial history. Understanding that history, and the theoretical framework developed to analyse it, is not optional for anyone serious about ethical AI.
Reflection question: Think of an AI system you interact with regularly. Whose experiences do you think were centred in its design? Whose might have been marginal or invisible?