The increasing popularity of artificial intelligence is leading to fears that it will replace workers. But the U.S. Equal Opportunity Commission is also looking at this technology to ensure it does not foster employment discrimination.
EEOC review
Last month, the EEOC conducted an online public hearing on employment discrimination in AI. It took testimony from computer scientist, civil rights advocates, and employer representatives. The EEOC Commissioner hopes that existing employment law, such as Title VII of the Civil Rights Act, can apply to AI and employment discrimination issues.
New York City has been among the first jurisdictions to address how employers can utilize automated employment decision technology. Starting on Jan. 1, New York City requires a bias audit before this type of technology may be used. Companies must also provide notice to job applicants and employees before using these methods.
Abuse
Employers should use variables that are highly predictive or corelate with successful job performance. But companies often use proxies for discriminatory preferences even if they are not screening for race, gender, sexual orientation, or other protected classes. Automated hiring programs are advertised as a way to replicate a company’s best worker which is a slogan that actually replicates bias.
A system that relies on criminal records and credit history for background checks may discriminate against Native Americans, or other workers of color because of racial profiling and redlining. Zip codes or college education can be used as race-based proxies.
Systems using certain names and sports, such as lacrosse, may also prioritize Caucasian and male workers. Social Security records indicate that certain names are associated with these groups.
Attorneys assist workers who face discrimination. They can help assure that their rights are protected under federal and New York laws.