Please ensure Javascript is enabled for purposes of website accessibility
Home / News / Reporting on AI hiring tools requested

Reporting on AI hiring tools requested

The Office of Federal Contract Compliance Programs (OFCCP) has issued a new requirement for government contractors to disclose their use of recruitment and hiring technology, including the use of artificial intelligence (AI).

The updated requirement now asks for information on “artificial intelligence, algorithms, automated systems, or other technology-based selection procedures.”

The OFCCP has reasoned that “use of these technology-based selection procedures may lead to instances of screening or selection bias such as assigning lower ratings to minority or women candidates in a screening process. Individuals with disabilities are also at risk of exclusion due to these tools.”

In establishing this requirement, the OFCCP intends to ensure that “contractors are evaluating whether their selection procedures, including the use of AI tools, are creating barriers to equal employment opportunity.”

This is the first time that the OFCCP has requested information on hiring and recruitment technology as part of its scheduling letter. Critics say that the language is ambiguous and could encompass any number of tools used in hiring, leading to onerous compliance reporting.

Employers should be aware that the OFCCP has a memorandum of understanding with the Equal Employment Opportunity Commission (EEOC), which means that information submissions may also be accessed by the EEOC. Both agencies have expressed concern over the potential for AI to introduce bias into employment decisions, and the EEOC recently settled its first AI bias lawsuit.

The OFCCP scheduling letter applies to supply and service compliance evaluations scheduled on or after August 24, 2023. Contractors should ensure that they are familiar with the requirements and, furthermore, exercise caution when adopting hiring and recruitment tools.

Contractors are advised to evaluate their tools to determine how AI or algorithms are being used and take steps to minimize risk. Periodic audits should be used to test for disparate impact.