The Electronic Privacy Information Center, a public interest research center based in Washington, D.C., recently asked the Federal Trade Commission to investigate HireVue, a recruiting company based in Utah that purports to evaluate a job applicant’s job qualifications through online “video interview” and/or “game-based challenge.”

According to its web site, HireVue has more than 700 customers worldwide including over one-third of the Fortune 100 and such leading brands such as Unilever, Hilton, JP Morgan Chase, Delta Air Lines, Vodafone, Carnival Cruise Line, and Goldman Sachs. The company states it has hosted more than ten million on-demand interviews and one million assessments.

The EPIC complaint follows a wave of lawsuits in recent years charging that employers are using software algorithms to discriminate against older workers by targeting internet job advertisements exclusively to younger workers.

HireVue uses a proprietary algorithm to assess “tens of thousands of data points” from candidate video interviews, including the candidate’s “intonation,” “inflection” and “emotions.” These and other data points are input into “predictive algorithms” that compare candidates with a company’s top performers.

HireVue states its video-based algorithmic assessments provide “excellent insight into attributes like social intelligence (interpersonal skills), communication skills, personality traits, and overall job aptitude.”

EPIC states HireVue’s “business practices produce results that are biased, unprovable and not replicable.”

The EPIC complaint focuses upon potential bias against women and minorities but the concept of using video game-based assessments seems particularly suspect with respect to older workers.

The  average age of video game players is in the mid-30s.  Many older workers do not play video games and it is likely that fewer older women than men have done so.  Even if they have, their response would likely be slower than a skilled young gamer.

Loren Larsen, Chief Technology Officer at HireVue, states in a press release that game-based assessments measure ‘emotional IQ” and “build upon the understanding of cognitive ability, enabling hiring managers to understand more about a candidate’s abilities.”

EPIC says HireVue “collects facial data from job candidates to evaluate expressions of emotion and personality” It says so-called “facial action units” make up to 29 percent of a candidate’s score.

But how does HireVue’s algorithm assess overweight candidates, those who suffer from depression or non-native English speakers? What about candidates with autism who tend to look at people’s mouths and avoid direct eye contact?

HireVue issued a press release in September announcing it has partnered with Integrate Autism Employment Advisors, a nonprofit organization that helps companies recruit “qualified professionals on the autism spectrum.” The 600 candidates in Integrate Autism’s network will be coached on how to use HireVue’s platform more “efficiently and effectively.” Would such an effort be necessary if HireVue’s algorithm didn’t penalize autistic candidates.

EPIC states the University of Maryland Robert H. Smith School of Business tells students preparing for AI interviews, “Robots compare you against existing success stories; they don’t look for out of-the-box candidates.”

EPIC notes that Amazon abandoned an AI recruiting tool after learning that historical employment data preferred male candidates over women.  In that case, the system penalized resumes that included the word “women’s” and the names of all-women’s colleges. EPIC contends the only way to really tell if a hiring algorithm is non-biased is to  “turn it off and hire someone using other criteria to figure out if they meet or fall short of the same performance standards.”

The bottom line is that HireVue’s algorithm is secret so the public has no way of knowing whether it discriminates against older workers, women, the disabled and minorities, etc. EPIC says HireVue refuses to provide job applicants with their assessment score.

EPIC charges that HireVue violates the Federal Trade Communications Act because it lacks a  “reasonable basis” to support its claims and cannot be held accountable for the proper functioning of its secret algorithmic assessments. EPIC also disputes HireVue’s contention that it does not use facial recognition technology for identity recognition purposes, which has been found to be a violation of the FTC Act.

EPIC alleges HireVue’s platform also fails to meet minimal standards for AI-based decision-making set out in AI Principles approved by the 36-member countries of the Organisation for Economic Co-Operation and Development (OECD), including the U.S. These principles say everyone has a right to know the basis for an AI decision that concerns them, AI systems should be deployed only after an adequate evaluation of its risks and institutions must ensure that AI systems do not reflect unfair bias.

Source : https://www.forbes.com/sites/patriciagbarnes/2019/11/10/artificial-intelligence-poses-new-threat-to-equal-employment-opportunity/?ss=ai#544eed306488