AI in the Racial Wealth Gap

This qualitative report is based on interviews with people of color who believe they have been impacted by AI decision-making systems in their pursuit of jobs, lending, and housing. Key takeaways include:

1. AI is scaling existing patterns of racial discrimination. Systems trained on historically biased data reproduce those biases across millions of decisions simultaneously.
2. The harm is not domain-specific. Housing, employment, and lending are connected by shared AI-powered infrastructure. Exclusion in one domain actively damages access in the others.
3. Communities of color are disproportionately exposed to AI’s impacts on wealth. 91% of Black Americans see AI as a job threat; 24% of Black workers are in roles with 75%+ automation potential; Black mortgage applicants are 80% more likely to be denied.
4. AI hardens the structural barriers faced by first-generation wealth-builders. When AI has been given the authority to make life-altering decisions, it cannot take context into account in the same way that a human decision-maker could.
5. People cannot fight what they cannot see. The transparency deficit means most participants had no idea AI was involved in decisions that shaped their financial lives — and no path to appeal.
6. Generational wealth feels unattainable to many. The compounding weight of algorithmic barriers makes generational wealth feel structurally out of reach, not just personally difficult.
7. AI could help close the racial wealth gap, with safeguards. With guardrails and equitable development and deployment, AI could be a beneficial tool to support economic mobility and increase generational wealth.