Black researchers are underrepresented where it matters most

The gap isn't just in AI broadly. Black researchers remain underrepresented in safety (alignment, evals, and testing), security, and governance, the fields where decisions get made about how powerful systems behave, what risks matter, and which policies shape AI's future.

Talent exists. Pathways are what's broken.

High-potential Black researchers are entering AI, but being funneled into capabilities roles rather than safety, security, and governance work. The talent exists. What's missing are clear, supported pathways into alignment research, evaluations, security roles, and the policy positions that shape how AI is built and regulated.

Blind spots in safety, security, and governance have real consequences

When safety researchers, security professionals, and governance experts lack lived experience with algorithmic bias in criminal justice, discriminatory hiring, or predatory financial AI, those risks get underweighted in evals, red-teaming, and policy frameworks alike. Broader participation isn't just equitable. It produces better, more complete work across all three fields.

Read more about our call to action!