The Guidelines contain questions and answers on the use of artificial intelligence (AI) and algorithms in human resources’ selection process. Here is a selection of these Q and A.

What is an algorithm?

It is a set of instructions that a computer can follow to achieve a specific purpose. In the field of human resources, they allow employers to process data to evaluate, rank, and make other decisions about candidates and employees.

What is AII?

Congress has defined AI as a "machine-based system capable, for a given set of human-defined objectives, of making predictions, recommendations, or decisions that affect real or virtual environments." In the employment context, AI generally means that the developer relies in part on the computer's analysis of data to determine which criteria to use in making decisions.

Can an employer's use of an algorithmic decision-making tool be a "selection procedure"? 

Yes. According to the Guidelines, a "selection procedure" is any "measure, combination of measures or procedure" as the basis for an employment decision. 

Can employers evaluate using an algorithmic decision-making tool for adverse impact in the same way they evaluate more traditional selection procedures? 

Yes, they can assess whether a selection procedure has a negative impact on a particular protected group by testing whether the use of the procedure causes a selection rate for individuals in the group that is "substantially" lower than the selection rate for individuals in another group. 

Is an employer liable for using algorithmic decision-making tools even if the tools are designed or operated by another entity, such as a software vendor?

Yes, in many cases. Therefore, they should ask the provider whether steps have been taken to assess whether using the tool causes a substantially lower selection rate for persons with a characteristic protected by law.

What is the 'selection rate'?

The 'selection rate' refers to the proportion of candidates or aspirants hired, promoted, or otherwise selected.

What is the 'four-fifths rule'?

According to the Guidelines, it is a rule that states that one rate is substantially different from another if their ratio is less than four-fifths (or 80 per cent). 

Does compliance with the four-fifths rule ensure that a particular hiring process does not have a negative impact?

Some Tribunal has already stated that it is only sometimes appropriate. It is not a reasonable substitute for a statistical significance test. 

The Guidelines highly recommend following the regulatory scheme as set out in the Artificial Intelligence and Algorithmic Fairness Initiative, launched by EEOC in 2021 to ensure that the use of software, including AI, machine learning, and other emerging technologies used in hiring and other employment decisions comply with the federal civil rights laws that the EEOC enforces.

The Consumer Financial Protection Bureau, the Department of Justice's Civil Rights Division, EEOC, and the Federal Trade Commission recently released a joint statement on Enforcement Efforts Against Discrimination and Bias in Automated Systems.

In the US, several states and localities also introduced laws to regulate the use of AI in automated employment decision tools (California, District of Columbia, Illinois, Massachusetts, New Jersey, New York, and Vermont).