A congressional subcommittee held hearings this week examining how federal agencies are deploying artificial intelligence systems for decision-making in areas ranging from benefits administration to law enforcement.
Witnesses testified about both the potential efficiency gains and the risks of algorithmic bias in government AI systems. Several advocacy groups presented evidence of disparate impacts on minority communities from automated screening tools.
Committee members from both parties expressed interest in establishing mandatory transparency requirements and regular audits for AI systems used in federal decision-making processes.
