EEOC issues guidance on using AI in employment decisions

Employers can assess whether using algorithmic decision-making tools in employment decisions is having an adverse impact just as they would for more traditional selection procedures, the U.S Equal Employment Opportunity Commission said in guidance issued Thursday.

The 12-page EEOC guidance addresses questions employers may have on how to avoid causing disproportionally negative effects regarding race, color, religion, sex or national origin in violation of Title  VII of the Civil Rights Act of 1964 as a result of using these tools.

The guidance says an employer is responsible under Title VII for its use of algorithmic decision-making tools, even if they are administered by another entity, such as a software entity.

It says the four-fifths “rule of thumb” is useful in determining whether an employment selection rate for one group is substantially different from others.

An example is a situation where the selection rate for Black applications was 30% versus 60% for white. The EEOC, however, said this rule may be inappropriate in certain instances.

The guide says employers can take steps to reduce a tool’s impact, or use a different tool, if it finds the use of algorithmic decision-making would have an adverse impact.