As cities roll out new laws governing the use of A.I. in the workplace, more employers are taking measures to cover their backsides from potential lawsuits.

The risks are high, with so much still unknown about the rapidly evolving technology and its potential ramifications. A New York City law, which will go into effect next month, aims to protect job candidates from potential biases in recruiting processes utilizing A.I., and the Equal Employment Opportunity Commission recently reminded employers that they’re responsible for any discrimination in hiring, firing, or promotions that are a result of A.I., even if it’s the fault of a third-party vendor.

Michael Schmidt, a labor and employment attorney with Cozen O’Connor, says employers would be wise to establish robust policies and internal checks to ensure protection from legal risks, likening it to the early days of moderating social media use in the workplace. 

“You are applying a new platform and technology to traditional employment law issues,” says Schmidt. “We need to figure out how to apply the same risks and rewards when it comes to third-party intellectual property, harassment, discrimination, and accurate and good content.” 

He suggests that HR leaders begin by auditing the technology vendors they already employ in hiring.

“Very often, organizations don’t even know that third parties that they have assisting them with the recruitment process are using A.I. as part of that recruitment process,” says Schmidt. “The burden is on the employer to make sure and not just assume that the third party is compliant, but to really look into what the third party is doing and how they’re doing it.”

HR leaders should ask vendors what A.I. is used in the recruitment process and how data is used and gathered, he tells Fortune. Then, perhaps most importantly, evaluate how your operations comply with state and federal regulations.

Schmidt recommends breaking the process down into two steps. First, assess the nature of your workforce and the workplace, and ask what your company is trying to achieve with its A.I. policy. Is it security, antidiscrimination, or protection from plagiarism? Is it to evaluate whether the use of A.I. is warranted at all?

Once an organization decides whether to embrace or limit the use of A.I. and in what ways, Schmidt says leaders should then map out the particulars of who will be using it and to what extent. For example, he says employers might allow the use of generative A.I. in employees’ rote decision-making but not replace decision-making altogether.

“The takeaway of all of this is that A.I. really has many advantages for employers when used correctly and appropriately,” says Schmidt. “It’s just that employers need to stay on top of what the regulatory landscape looks like in the technology landscape to make sure that they are using it appropriately.”

Amber Burton
amber.burton@fortune.com
@amberbburton

Reporter's Notebook

The most compelling data, quotes, and insights from the field.

The tech sector trend of firing middle managers to create a flatter organization can be attributed to Elon Musk—or so says Meta CEO Mark Zuckerberg. 

"His actions led me, and I think a lot of other folks in the industry, to think about, ‘Hey, are we kind of doing this as much as we should?’” Zuckerberg said in a podcast interview last week. “My sense is that there were a lot of other people who thought that those were good changes but who may have been a little shy about doing them."

Around the Table

A round-up of the most important HR headlines, studies, podcasts, and long-reads.

- Digital nomads are growing disillusioned with their work-from-anywhere way of life and returning to a more traditional lifestyle. BBC

- LinkedIn plans to overhaul its platform to make hiring for skills over education easier for employers. Bloomberg

- Remote jobs are still present but becoming increasingly concentrated in the Northeast. Wall Street Journal

- In an all-hands meeting last week, Meta CEO Mark Zuckerberg promised to rebuild the company’s culture after recent layoffs. New York Times

- The gender pay gap has cost women $61 trillion since 1967, according to a report from the Center for American Progress.

Watercooler

Everything you need to know from Fortune.

Alphabet workers complain. Google’s employee union pushed back against management’s decision to include office attendance in performance reviews, calling the practice “ambiguous.” —Prarthana Prakash

Fool me once. ChatGPT fooled a lawyer into citing fake court cases in a legal document, raising questions about A.I.’s ability to fabricate documents in the workplace. —AP

The ol’ college try. Some 90% of the 2023 graduating class say college was worth the expense, according to a Harris Poll survey. —Tim Killeen, Will Johnson

Actually, come back. Y Combinator cofounder Paul Graham shared some musings on companies' remote work 180, tweeting, "I doubt things will go all the way back to the way they were before Covid, but it looks like they will go most of the way back."