AI

Silly, inefficient humans discuss AI at HR Transform

Experts, including EEOC Commissioner Keith Sonderling, discussed the ethical dilemma of AI.
article cover

Carlos Castilla/Getty Images

· 4 min read

Artificial intelligence (AI) is used more than ever across HR functions, including initial resume screening, employee engagement, and performance monitoring, but critics say it’s also perpetuating long-standing problems, from bias around language analysis to employee privacy issues.

There’s a flurry of conversation around using AI in HR functions because AI relies on existing datasets and humans with biases develop the technology. Bias in AI-powered hiring software has been identified in recent years, including in 2018 when Amazon shut down its AI recruiting tool because it found it was biased against women.

Policymakers are writing legislation to combat new problems arising from AI, including a new law in New York City that requires AI tech vendors to obtain and display third-party “bias audits” for transparency and accountability. And in February, Democrats in Congress introduced an updated version of a bill, the Algorithmic Accountability Act, which would hold tech companies accountable for algorithmic bias across sectors, from housing to employment.

The ethical and philosophical implications of using AI technology in HR were discussed at length during this month’s HR Transform conference in Las Vegas. Executives from Randstad, Microsoft, Sequoia, and others discussed the best ways to use AI. Here are the biggest takeaways.

Understanding the why behind data analysis will help teams choose the right solution, according to Rajamma Krishnamurthy, senior director, HR technology at Microsoft. She noted that teams must know what problem they’re trying to solve before choosing a solution, and regularly audit data to ensure bias isn’t being built into the system. ”People assume that the perfect mathematical equations are perfect,” Krishnamurthy said, and emphasized that when something goes wrong with AI, “there are only three things: data, data, data.”

EEOC commissioner says...Participating on a panel about AI and ethics, EEOC Commissioner Keith Sonderling argued that we all have inherent biases that can negatively impact the employment process, and HR professionals often turn to these new technologies in part to try and remove human bias.

HR is challenging. HR news doesn’t have to be.

HR Brew keeps you effective in the fast-changing business environment.

Sonderling urged employers to be cautious when choosing where to implement AI tech. “I would say the use of AI to evaluate and score video interview content is one area where I would say it’s extremely high-risk,” explained Sonderling. “If the computer can’t transcribe what they're saying because the candidate is from a different country and has a foreign accent, or the candidate is disabled, and they can’t actually account for what they’re saying, they may be ranked lower than somebody who speaks perfect English.”

​​And when it comes to collecting employee data, Amit Mohindra, founder and CEO of People Analytics Success, an advisory and education firm for HR professionals, argued that it’s better to err on the side of transparency. “I think you always have to be transparent about what you’re doing with the data, and how it’s going to influence [employees] and essentially sell them on what’s in it for them, which is better outcomes for talent programs, and better outcomes for the company.”

At the core of every new AI-enabled HR solution, employee and candidate civil rights must be protected, Sonderling said. He explained that while artificial intelligence “promises to remove the human bias from these applications by removing the human,”from functions like resume screening, “those computers and those programs need to understand some of the very significant civil-rights applications that my agency (EEOC) has to deal with.”

Zoom out. Feeling overwhelmed at the idea of using artificial intelligence? Sonderling emphasized, “it’s a lot of the same analysis you’ve been doing for r decades, just applying that to modern technology.”—KP

Do you work in HR or have information about your HR department we should know? Email [email protected] or DM @Kris10Parisi on Twitter. For completely confidential conversations, ask Kristen for her number on Signal.

HR is challenging. HR news doesn’t have to be.

HR Brew keeps you effective in the fast-changing business environment.