Pathology/Lab Coding Alert

Artificial Intelligence:

Use Caution With AI Staff Recruiting Tools

Avoid legal pitfalls.

Yet another transformative technology is seeping into the healthcare industry. In fact, you may already be using artificial intelligence (AI) in seemingly straightforward ways, such as doing an initial resume screening for hiring.

Caution: But the Equal Employment Opportunity Commission (EEOC) as well as AI stakeholders are warning of potential dangers. Whether you’re already neck-deep in using AI this way or just now considering it for your lab, keep reading to see what to look out for.

Get Perspective on AI Then and Now

AI has been used for years to simplify myriad processes. You encounter it when you use a search engine or use autocorrect on your phone or spellcheck on your computer. In his Senate appearance, OpenAI CEO Sam Altman described an individual with dyslexia who uses a specialized AI tool built on OpenAI technology to help him navigate email and other written communications for his job, including securing a $260,000 grant.

Now: Some human resources (HR) departments may use AI technology to scan resumes, flagging certain words or phrases so that resumes can be moved to the next round or be removed from the pool.

Understand Who is Advocating AI Regulation and Why

With AI technology progressing so rapidly, some federal agencies and other regulators are more closely scrutinizing its use. In fact, some of the developers of the more advanced AI technology are advocating for extensive regulation, including Altman, who appeared before the U.S. Senate to talk about some of the ways AI can benefit people, as well as its various dangers.

For example, the EEOC is warning employers that some of the algorithms that AI uses may be biased against some people. Employers may be inadvertently violating protections outlined in the Americans with Disabilities Act (ADA) or the Civil Rights Act of 1964 when using this technology in their hiring processes.

You may think your practice couldn’t be implicated in any AI scrutiny related to ADA or Title VII protections, but if you use the technology, you need to check in with your compliance team.

Brush Up on ADA, Title VII

The labor-relevant portions of the ADA prohibit employers and other entities from discriminating based on disability. Additionally, Title VII of the Civil Rights Act of 1964 prohibits employment discrimination based on race, color, religion, sex (including pregnancy, sexual orientation, and gender identity), or national origin.

Important: The EEOC says that employers can be held liable for AI biases affecting protected populations in the hiring process even if they did not personally design the AI software or its respective algorithms. This means employers should review how they or their vendors utilize AI technology in their hiring practices, says Kathryn Jones, attorney at Hall, Render, Killian, Heath & Lyman, in online analysis. Additionally, they should take advantage of the technical assistance the EEOC is providing for employers that use AI technology in their hiring process, so they can check — and correct — any disparate impacts on protected populations.

Understand the Implications; Recall Four-Fifths Rule

If you’re now wondering about your own lab’s compliance risk when using AI in the selection process, the EEOC suggests considering the four-fifths rule.

Four-fifths rule: In a technical assistance document, the EEOC mentions the “four-fifths rule,” which is how they’re checking to see whether the rate of selection is “substantially” different from one group to another.

Although the EEOC is careful to designate the four-fifths rule as a “rule of thumb” rather than a hard and fast metric, understanding the calculus is important to check your own compliance. “The four-fifths rule … is a general rule of thumb for determining whether the selection rate for one group is ‘substantially’ different than the selection rate of another group. The rule states that one rate is substantially different than another if their ratio is less than four-fifths (or 80%),” the EEOC says.

Note: “Courts have agreed that use of the four-fifths rule is not always appropriate, especially where it is not a reasonable substitute for a test of statistical significance. As a result, the EEOC might not consider compliance with the rule sufficient to show that a particular selection procedure is lawful under Title VII when the procedure is challenged in a charge of discrimination,” the EEOC says.

While the four-fifths rule may not constitute statistical significance, the EEOC still recommends that practice managers ask AI vendors whether they used the four-fifths rule when evaluating whether their product might have an adverse impact on a Title VII characteristic.

Remember: Anyone can file a charge if they feel like they’ve experienced discrimination related to employment. “A discrimination charge is an applicant’s or employee’s statement alleging that an employer engaged in employment discrimination and asking the EEOC to help find a remedy under the EEO laws,” the EEOC says. The EEOC may then begin an investigation and can also assist in confidential mediation to pursue resolution.

Takeaway: While AI is more than capable of helping people with disabilities, AI can also be used to screen and eliminate job applicants or candidates who may need accommodations — a violation of the ADA. If your practice is using, or considering using, AI to help with recruitment, keep this important potential EEOC violation risk in mind.