This paper examines the history of data-driven housing discrimination in the United States, beginning with the practice of “redlining” in the 1930s and closing by examining several current “smart city” practices for algorithmic housing allocation. “Redlining” never truly ended as a practice; rather, it shifted forms as housing allocation became increasingly privatized in the late 20th century. Spatial racism has always been data-driven in the United States, but contemporary algorithmic sorting tools increasingly allow landlords and mortgage lenders to easily advance claims of fairness and neutrality while reproducing discriminatory outcomes. Such tools and outcomes reinforce decades of systemic racism in the United States’ housing market, particularly anti-Black racism.