As much excitement as there has been regarding the possibilities of artificial intelligence in numerous fields – much of it justified – the technology has also provoked a certain degree of skepticism and trepidation. In recent months, these issues have become a concern in onboarding and staffing.
State legislators in Illinois passed a law to address the use of AI in the hiring process that went into effect right at the start of 2020. No other states have yet followed Illinois's example. However, the continuing nature of conversations regarding AI use (and possible abuse) means that the question of whether those governments – or perhaps the U.S. government as a whole – move to introduce similar legislation is more likely a matter of "when" than "if." Hiring managers and other HR leaders who are considering adopting such tools would do well to consider the Illinois law and why it was passed, as well as the larger implications at play that could ultimately affect countless organizations.
Basics of the law
According to the National Law Review, all Illinois businesses that conduct video interviews and use AI to analyze the resulting footage must inform job applicants that they're engaging in those practices under the terms of the Artificial Intelligence Video Interview Act. Such companies are also required to explain how their AI platform of choice works and what details it examines to reach its conclusions, and lastly hold the obligation to receive the applicant's consent for AI video analysis before the interview takes place. This law applies to both live interviews conducted via conferencing tools like Skype or Zoom and formats in which candidates receive pre-recorded questions and then make videos of their responses.
Further stipulations apply after prospective employees and recruiters have had their virtual conversations. According to the Chicago Tribune, candidates can opt out of video interviewing (though a replacement format isn't specified under the law, leaving room for employers to write off such individuals), or ask for their recordings to be destroyed. Companies are also only allowed to share the footage with those "whose expertise or technology" is required to make a hiring decision – e.g., other members of HR, those who would supervise the new hire or any third-party consultants commissioned to assist in such matters.
Illinois Representative and AIVIA co-sponsor Jaime Andrade Jr., who told the Tribune he was working on a follow-up to mandate alternate interview formats, also described the issue at stake succinctly to the news provider: "[The software] is only as good as the data it's given."
Precedents and successors
To those only hearing about it now, the AIVIA may seem almost shockingly prescient for going into effect Jan. 1 of this year – months before COVID-19 became a full-scale pandemic – now that a vast majority of companies are only hiring virtually for the indefinite future. But numerous companies had been using video interviews (and, in some cases, AI) well before the AIVIA's passage.
In fact, the Biometric Information Privacy Act, another Illinois law that addresses similar concerns regarding biometric scans of individuals' faces and fingerprints, predates the AIVIA by almost 12 years. Three other states now have similar laws – Texas, Washington and California – and similar legislation centered around consumer privacy has been considered in Massachusetts, Arizona and Florida.
The heart of the matter
A November 2019 debate prompted by one particular AI-based video interview platform sums up the central issues surrounding such technologies. The Washington Post reported that HireVue, a system analyzing job applicants' word choices, voices and facial expressions to determine their suitability for a given position using a vague scoring system, generated criticism from AI experts and prompted the Electronic Privacy Information Center to file a formal Federal Trade Commission complaint regarding deceptive practices.
"It's a profoundly disturbing development that we have proprietary technology that claims to differentiate between a productive worker and a worker who isn't fit, based on their facial movements, their tone of voice, their mannerisms," AI Now Institute co-founder Meredith Whitaker told the Post. "It's pseudoscience. It's a license to discriminate."
HireVue CTO Loren Larsen vehemently opposed the platform's critics, saying they didn't understand AI and that it had no more (or fewer) flaws than traditional hiring practices.
Larsen's product isn't unique; it earned massive attention due to its use by high-profile corporations including Unilever and Hilton (and the Post's coverage). Its competitors likely use similarly controversial algorithms. Employers must weigh the benefits such technologies can bring to the hiring process, particularly in a largely remote environment, against the possibility of compliance issues down the road if other states or the federal government adopt laws like the AIVIA.