Digital Civil and Human Rights Act of 2022
A BILL
To establish prohibitions on the use of automated systems in a discriminatory manner, and for other purposes.
Sec. 2 Prohibition of discriminatory use of automated systems
“(f) It shall be an unlawful practice for a place of public accommodation to use in its operations affecting commerce any automated decision system (as such term is defined in section 701) that has a disparate impact on the basis of race, color, religion, sex, or national origin.”
“(o) It shall be an unlawful employment practice for a respondent, in connection with the selection or referral of applicants or candidates for employment or promotion, to use any automated decision system that has a disparate impact on the basis of race, color, religion, sex, or national origin.”
“(o) The term “automated decision system” means any system, software, or process (including one derived from machine learning, statistics, or other data processing or artificial intelligence techniques and excluding passive computing infrastructure) that uses computation, the result of which serves as a basis for a decision or judgment.
“(p) The term “court of the United States” has the meaning given such term in section 451 of title 28, United States Code.”
Sec. 3 Prohibition of bias in use of automated systems by the department of defense
“130l. Prohibition of use of automated systems with bias
“(a) In general—None of the funds authorized to be appropriated or otherwise made available for any fiscal year for the Department of Defense may be used for the use of any automated decision system with any adverse distinction based on race, religion, sex, health, age, or any other similar criteria.
“(b) Publicly available repository—Not later than one year after the effective date of this section, the Secretary of Defense shall maintain a site on the public website of the Department of Defense that contains a bias datasheet for each automated decision system used by the Department of Defense with the potential for an adverse distinction on the bases established under subsection (a).
“(c) Definitions—In this section:
“(1) The term “automated decision system” has the meaning given such term in section 701 of the Civil Rights Act of 1964.
“(2) The term “adverse distinction” has the meaning given such term through the application of the Geneva Conventions and customary international law.
“(3) The term “bias datasheet” means a datasheet conforming to the standard published pursuant to section 22B of the National Institute of Standards and Technology Act and includes an addendum related to international law.”
Sec. 4 Report on intelligence community use of automated decision systems with bias
Sec. 5 Establishment of datasheet standards on automated decision system bias
“22B. Automated decision system bias datasheets
“(a) In general—The Director of the Institute, in consultation with such researchers, private sector industry experts, legal experts, and civil rights experts as the Director considers appropriate, shall develop, publish, and maintain standards for reporting bias in an automated decision system.
“(b) Elements—The standards described in subsection (a) shall require the following:
“(1) An assessment of bias on the basis of race, national origin, color, religion, disability, sex, sexual orientation, or gender identity.
“(2) The source, size, and characteristics of any dataset used as the basis for the automated decision system.
“(3) The content is accessible and machine readable in accordance with the 21st Century Integrated Digital Experience Act (44 U.S.C. 3501 note).
“(4) Any other requirements as determined by the Director.
“(c) Definitions—In this section the term “automated decision system” has the meaning given such term in section 701 of the Civil Rights Act of 1964.”