Using AI in Hiring or Recruiting: What You Need To Know

Woman in business attire on a video interview meeting on her laptop.

 

The rapid rise of free generative artificial intelligence tools has workplaces abuzz with what comes next. But if you work in human resources or your business has used software to find job candidates, you may be more concerned with the immediate ramifications of AI.

The HR news site HR Dive reports that there are 50 more applicants per job opening this year than last year. And for some industries, the number is much higher. Sorting through resumes manually can be a pain, especially when you have several open positions. Modern hiring software can free you from this tedious task. It can scan for keywords and experiences, and even analyze social media accounts to find the best match for the role.

Some AI tools take it even further, allowing candidates to answer interview questions on recorded video. The tools then analyze their responses, including facial expressions and tone of voice, for characteristics such as “conscientiousness” or “openness.”

The legal implications of using AI for hiring and recruiting

Though AI tools are helpful, they are imperfect. Video hiring software has been known to assign candidates different scores based on what they’re wearing, such as glasses or a headscarf. These programs have even rated women lower in interviews for tech positions.

While this discrimination may not be deliberate, it is illegal. Federal law prohibits employers from discriminating against candidates based on protected traits, such as race, color, religion, sex, national origin, age, disability and genetic information. If your organization is using software that discriminates against candidates based on any of these factors, you could face legal trouble.

If you decide to use AI for hiring and recruiting, consider screening your results for “disparate impact.” Disparate impact is when a rule or process seems innocent or neutral but actually results in discrimination against a group of people. A rule or process can be discriminatory even if discrimination wasn’t the intent.

For example, if your AI tool consistently yields better results for men than women, it’s producing a disparate impact. You must modify the tool to close that gap.

AI legislation to pay attention to

The federal government has not issued concrete guidance on what you should do to avoid AI bias in hiring and recruiting. However, decades of precedent have shown that actions or processes that produce a disparate impact are illegal. Also, while the federal government may be slow to issue guidance, several states and localities are moving quickly.

What should you do to avoid discrimination in hiring and recruiting?

If any part of your hiring or recruiting processes involves software, consult with legal counsel and us. We can answer your questions, help you audit your processes and keep you updated on legal developments.