Letter to DOJ re: AI Executive Order
Upturn, Center for Democracy & Technology, Lawyers’ Committee for Civil Rights Under Law, & seven groups
LetterRE: Comprehensive Use of Civil Rights Authorities to Prevent and Combat
Algorithmic Discrimination
Dear Attorney General Garland and Assistant Attorney General Clarke,
We, the undersigned groups, write to you with specific recommendations on
implementing the recently-signed Executive Order on Safe, Secure, and Trustworthy
Development and Use of Artificial Intelligence (EO 14110). The EO directs the Department
of Justice’s Civil Rights Division (DOJ CRT) to convene, within 90 days of the date of the
Executive Order, a meeting of the heads of Federal civil rights offices to “discuss
comprehensive use of their respective authorities” to “prevent and address discrimination
in the use of automated systems, including algorithmic discrimination.” It also directs you
to “increase coordination between the Department of Justice’s Civil Rights Division and
Federal civil rights offices concerning issues related to AI and algorithmic discrimination,” and to “develop, as appropriate, additional training, technical assistance, guidance, or
other resources.”
We commend the work that DOJ CRT and other civil rights enforcement agencies have
done so far in combating algorithmic discrimination, and for quickly convening in response to EO 14110. In particular, we are grateful that the participants at the convening “pledged to continue collaboration” to combat algorithmic discrimination and also “agreed to partner on external stakeholder engagement around their collective efforts to
advance equity and civil rights in AI.”
However, more can and must be done. Two years ago, Assistant Attorney General Clarke noted that the DOJ CRT was “reviewing whether guidance on algorithmic fairness and the use of AI may be necessary and effective.” We believe that such guidance — and more — is necessary, especially in light of the Executive Order and the urgency due to the harms of algorithmic discrimination. In particular, the undersigned groups urge you to:
Develop and issue an Interagency Policy Statement on algorithmic
discrimination. While the CFPB, DOJ CRT, EEOC, and FTC jointly stated that
existing legal authorities apply to the use of automated systems, more guidance is
necessary. An Interagency Policy Statement would describe the general principles
agencies will consider in determining whether unlawful algorithmic discrimination
exists and provide a foundation for future regulatory and enforcement actions.Alongside civil rights enforcement agencies, establish an interagency working
group to develop and expand the federal government’s own
anti-discrimination testing capabilities to uncover algorithmic
discrimination. A dedicated interagency working group should be charged with
developing and expanding anti-discrimination testing capabilities, assistance on
enforcement cases, and other efforts to combat algorithmic discrimination.Pursue opportunities to require covered entities to perform regular
anti-discrimination testing of their systems and regularly search for less
discriminatory algorithms. Existing authorities support these approaches.
Beyond an Interagency Policy Statement, agencies should take steps to clarify
expectations on the private sector to proactively combat algorithmic
discrimination.Urge civil rights enforcement agencies to require covered entities to collect
demographic information for anti-discrimination purposes. Efforts to combat algorithmic discrimination will be most effective when covered entities are
required to collect certain demographic information for anti-discrimination testing
purposes.Hold monthly meetings to provide external stakeholders with regular updates
about progress on AI EO implementation. DOJ CRT should also host an annual
workshop focused on research and methodologies aimed at combating algorithmic
discrimination, similar to the role the FTC’s PrivacyCon plays for consumer privacy
and data security.
The DOJ CRT, alongside its partner civil rights enforcement agencies, has the responsibility to launch a landmark effort to ensure that covered entities address algorithmic discrimination in civil rights areas. Each of the recommendations we set forth will require civil rights enforcement agencies to ensure that sufficient staffing resources are in place. We would be pleased to discuss the proposals in this memo in more detail in the weeks and months ahead.
1
Executive Order 14110, Section 7(a)(ii), “Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence,” 88 Fed. Reg. 75191 (Nov. 1, 2023), https://www.federalregister.gov/documents/2023/11/01/2023-24283/safe-secure-and-trustworthy-develop ment-and-use-of-artificial-intelligence.
1
Executive Order 14110, Section 7(a)(ii).
1
Id.
1
Department of Justice, “Readout of Justice Department’sInteragency Convening on Advancing Equity in Artificial Intelligence,” (Jan. 11, 2024), https://www.justice.gov/opa/pr/readout-justice-departments-interagency-convening-advancing-equity-artificial-intelligence.
1
Id.
1
Assistant Attorney General Kristen Clarke, “Keynote on AI and Civil Rights for the Department of Commerce’s National Telecommunications and Information Administration’s Virtual Listening Session,”(Dec. 14, 2021), https://www.justice.gov/opa/speech/assistant-attorney-general-kristen-clarke-delivers-keynote-ai-and-civil-rights-department.