The rise of easily accessible artificial intelligence (AI) tools is changing every industry, and one of the many new applications of AI is in making management decisions.
While AI has promise to improve efficiency and productivity, it also carries significant risks for employees. In particular, concerns have been raised about AI algorithms making discriminatory employment decisions, creating unsafe working conditions, and otherwise infringing on workers’ rights.
Several proposed pieces of federal legislation seek to address the risks posed by AI and other new technologies in the workplace. One, the No Robot Bosses Act, would prohibit employers from using automated decision systems to make employment decisions, including hiring, disciplinary actions, and firing. Another, the Stop Spying Bosses Act, would prohibit the use of electronic surveillance for certain purposes, including monitoring employees’ health, keeping tabs on off-duty workers, and interfering with union organizing.
“Working people MUST have a voice in the creation, implementation and regulation of technology,” said the AFL-CIO in support of the No Robot Bosses Act.
The role of technology in employment discrimination
While it’s impossible for an algorithm to be consciously bigoted, it’s absolutely possible for an AI to reach discriminatory employment decisions. AI hiring tools are developed using training data, which may in turn introduce human biases or historical inequality. In addition, certain groups may be over- or under-represented in the training data, which can inadvertently train the AI to make decisions biased for or against those groups.
For example, automated tools that scan job applicants’ resumes may develop a bias toward keywords that are more frequently found on men’s resumes than women’s. AI hiring tools that analyze applicants’ faces and speech patterns in video interviews may perpetuate bias on the basis of race or ethnicity. And when the hiring process actually involves interactive AI tools such as “chatbots,” those tools may create barriers for older workers or those who are less comfortable with technology (even if using technology is not a significant part of the job itself).
This is not just a hypothetical concern. When Amazon started using AI in hiring decisions, the tech giant immediately found that their AI tool discriminated against women. And earlier this month, the Equal Employment Opportunity Commission (EEOC) settled its first-ever AI hiring discrimination case, recovering $365,000 for a group of job-seekers.
AI bias in hiring is perhaps the easiest aspect of this problem to measure, but AI tools can also introduce or reinforce bias in promotions, raises, layoffs, and other employment decisions. And because of the way machine learning works, this has the potential to become a vicious cycle: the more AI makes biased decisions, the more that bias will become entrenched in the training data for the next generation of AI tools.
“We were just following the algorithm” is not an excuse
It’s important for employees to understand that as far as anti-discrimination laws are concerned, ignorance is not an excuse. To have a case, you don’t need to prove that your employer intended to discriminate on the basis of sex, race, religion, disability, age, or another protected characteristic. You only need to prove that their policies had a discriminatory effect on your employment.
As legislators grapple with how to update our laws to meet the changing employment landscape, employees need to know their rights and take proactive steps to protect them. If you believe you have been illegally discriminated against at work, contact us today for a free case evaluation with an experienced employment law attorney at Klie Law Offices.