Two Democratic Senators want OMB to direct agencies that use, fund, and procure AI for consequential decisions to establish civil rights offices. Image: Tada Images/Shutterstock.com
By: FEDweek StaffTwo leading Democratic senators have urged the Biden administration to expand controls over federal agency use of artificial intelligence, in particular to guard against bias in AI-assisted decisions.
“Without new protections, today’s supercharged, AI-powered algorithms risk reinforcing and magnifying the discrimination that marginalized communities already experience due to poorly-trained and -tested algorithms. The stakes — and harms — are especially high where entities use algorithms to make ‘consequential decisions,’ such as an individual’s application for a job, their treatment at a hospital, their admission to an educational institution, or their qualification for a mortgage,” Senate Democratic leader Chuck Shumer of New York, and Sen. Edward Markey, D-Mass., said in a letter to OMB.
They said that the administration’s 2023 executive order on agency use of AI represents a “good step forward” by, for example, creating the position of chief AI officer in agencies responsible for assessing potential bias and mitigating it. However, “OMB should also ensure that the CAIOs have the resources and expertise to address civil rights harm created by AI-driven algorithms. Additionally, OMB should direct agencies that use, fund, and procure AI for consequential decisions — defined broadly — to establish civil rights offices, if they do not already have one,” they wrote.
Such offices “should be staffed with technologists and experts in algorithmic discrimination whose job responsibilities include mitigating algorithmic bias and discrimination and facilitating proactive and ongoing outreach to civil rights stakeholders and affected populations,” they wrote.
Other recommendations included: enduring that the Justice Department “is effectively coordinating civil rights enforcement on AI among federal agencies”; setting standards for recipients of federal funds or contracts to identify and eliminate bias in algorithms they use; allow individuals to opt out of AI-assisted decisions and request human decision-making; and establish standards for enforcing prohibitions on bias in AI-powered decisions.
Key Bills Advancing, but No Path to Avoid Shutdown Apparent
TSP Adds Detail to Upcoming Roth Conversion Feature
White House to Issue Rules on RIF, Disciplinary Policy Changes
DoD Announces Civilian Volunteer Detail in Support of Immigration Enforcement
See also,
How Do Age and Years of Service Impact My Federal Retirement
The Best Ages for Federal Employees to Retire
How to Challenge a Federal Reduction in Force (RIF) in 2025
Should I be Shooting for a $1M TSP Balance? Depends…
