Australian EnterpriseAI Index
Back to incident database
CriticalGovernment23 May 2016

COMPAS Recidivism Algorithm Showed Racial Bias in US Courts

Northpointe (Equivant)

What happened

ProPublica's investigation found that Black defendants were nearly twice as likely to be falsely flagged as future criminals compared to white defendants. The tool was being used by judges to inform bail, sentencing, and parole decisions.

Root cause

Algorithmic model trained on historically biased criminal justice data; no independent bias audit prior to deployment in high-stakes judicial contexts.

Architectural failure

No bias detection or demographic parity testing; opaque model prevented scrutiny; deployed in high-risk decision contexts without human override requirements or explainability mechanisms.

Outcome

Ongoing legal challenges across multiple US states. Wisconsin Supreme Court ruled COMPAS scores could be used but judges must be cautioned. Prompted significant academic research and policy debate.

Architectural Failure Patterns

These pattern categories on aipatterns.com.au describe the systemic failure modes this incident exhibited.

Cite this incident

https://corporateai.com.au/incidents/compas-racial-bias-2016