Q&A: A Guidebook for Navigating Workers’ AI Pay Searches

“How much should I be getting paid?”

Employees and candidates are increasingly turning to online artificial intelligence (AI) tools with this question — and bringing the responses those tools generate to their supervisors or job interviews.

Workspan Daily (WD) spoke with Ben Eubanks, the chief research officer for Lighthouse Research & Advisory and the author of “Artificial Intelligence for HR: Use AI to Support and Develop a Successful Workforce,” about how employers, total rewards (TR) professionals and managers should frame and respond to this growing reality.

Ben Eubanks, chief research officer, Lighthouse Research & Advisory

Access a related Workspan Daily article on this subject:

• What Can You Do When Workers Use AI to Inform Their Pay Expectations?

Access a bonus Workspan Daily Plus+ article on this subject: 

• Workers Use AI for Pay Data. How You React and Respond Is Critical. 

WD: What are the pros and cons of employees and job seekers using AI to determine their workforce worth and generate pay expectations?

Eubanks: There’s a lot of data to support pay transparency. It creates more equitable outcomes for people. And, honestly, it just makes things easier.

When I was an HR leader, we had a hiring manager that would wait until it was uncomfortably, awkwardly late in the hiring process, because he didn’t want to be the first one to talk about pay. It was always unpleasant, and many times you’d find, “We’re so wildly off that this person is dropping out, and they’ve got a sour taste in their mouth about the company.” So, transparency is a good thing overall.

The cons: If they’re using AI to do this, AI is not always a reputable source. We ran an experiment recently looking at nurses in Dallas as an example, and said, “What are the pay expectations?” We used the same AI tool — I won’t call it out — but we asked it, over the course of 10 minutes, three different threads, three different prompts, and it gave us three different ranges in that time.

In my test, I’m being much more prescriptive and specific than a candidate is probably going to be, so their answers are going to be even more varied. Chances are they’re going to get answers that are higher, and then they’re going to show up and expect higher pay. They’re going to think they’re being mistreated, and the employer is then on the defensive, trying to respond to that.

WD: How might misinformation and misunderstanding muddy the TR value proposition?

Eubanks: It’s very hard to tell an employee that they’re wrong. It’s hard to do that in a way that’s respectful and in a way that salvages that relationship. But AI is pulling from sources like Reddit, where it found one person who said, “I make $100,000 as project manager.” And it doesn’t tell you, “I live in the Bay Area, and I’ve done this for 10 years.”

Then someone else Googles or uses AI to say, “What should my salary be as a project manager?” and it comes back and says, “Some project managers make $100,000 a year.” That person is going to show up at the office expecting that, and they don’t realize it’s a sample size of one.

The company is already on the back foot, trying to say, “No, no, we do have competitive pay, but here’s why you’re wrong.” It’s really hard to tell someone that. A hiring manager is more likely than a recruiter or HR team member to be the one having conversations about pay with candidates. And if that’s the case, they’re the ones that are typically not trained, not prepared and not aware of the company’s pay structures and how things are set up. They’re not equipped to have that conversation, and that creates misinformation and misunderstandings.

WD: How can TR pros shore up any shortcomings in the HR process or throughout the organization?

Eubanks: When we survey employers, and we look at the ones who are doing pay transparency well, one of the things that makes that group stand out is they aren’t just saying, “We’re sharing our pay ranges and job postings externally.” But they’re also thinking internally, “What can we share?”

That doesn’t mean they have to open-source the entire playbook on comp and what everybody makes, because that can be very strange for some cultures and some companies. But maybe they say, “These are the three main job levels in our company, and to get from a 1 to a 2, here’s what it takes; or from a 2 to a 3, here’s what it takes.” They make it more transparent and understandable, and tie it to pay. When they do that, it creates more trust. It creates more transparency. It creates a perceived level of connection with the company.

WD: How should TR pros and companies be thinking about this?

Eubanks: You need to be very clear internally what it takes to get your compensation changed in your company. If I asked an average employee walking down the hallway or someone who’s dialing in on a Zoom call, “Hey, before we get to this, I just want to ask you a quick question. Do you know how your pay is set and what it takes for you to move it to the next level?” I would say that 90% of employees could not answer that question, across the entire workforce population out there. And if that’s the case, how can we expect them to feel confident that the employer has their back when it comes to pay decisions? It’s silly to even expect that. So, if we want that, we have to be sharing.

I had a chance to talk to an HR leader about this, and she said, “Listen, I have to confess, I went too far. We shared more than we probably should have because we weren’t ready for it. And, we had people that were asking questions, and it got really sticky in some situations. We learned our lesson that we have to be guarded in what we share but be as open as we can. Always be pushing the boundaries of what’s comfortable. If it’s completely comfortable, you’re probably not sharing enough.”

That’s the approach I think most organizations should be thinking about. What else can we share? How can we default to transparency versus defaulting to the lack thereof?

WD: What might the next steps look like?

Eubanks: We need to equip our hiring managers to have that conversation. We need to make sure our recruiters are equipped to have that conversation. And, you don’t just send out an email. I would actually do some sort of training — talking about how your pay works overall and steering someone back on the path who shows up with a ChatGPT-generated pay rate that’s mythological.

WD: What is the hopeful result of taking steps to address workers’ use of AI tools in driving salary expectations (taking into account the knowledge that workers likely aren’t going to stop using these tools), and the gaps in worker trust and knowledge about internal pay decisions?

Eubanks: People just want to be respected. They want to know that someone actually does care about them. Pay is one of the things that we use to do that as a company. In their head, if they show up and they expect X and we offer them half of X, they’re already thinking, “Does this company respect me? Do they care about me?” It’s possible to change them over and make them think so again, but it’s not easy. It doesn’t happen by accident. It doesn’t happen by default. It has to be intentional.

Start the conversation with, “Listen, we really respect you as a candidate. We appreciate all the value you bring. We completely agree that you’d be serving a critical role if you take this job with us.”

Or if it’s an employee, have that conversation varied to the employee side. “We respect you. We appreciate you. We sincerely thank you for the effort and the ideas you bring to work every single day. Now, let me explain to you where this data comes from.”

Talk about some of the data sources and then you can — to the degree you want and the degree you can — share what your company is doing around pay, and be transparent about that. The more you can share, to the degree that it’s allowed within your company’s rules and policies and within your own culture as an organization, the better.

WD: Is there anything else you’d like to add?

Eubanks: People show up thinking about value. We’ve talked purely about comp. If you have a really rich benefits package, or you do profit sharing, or the company has a wonderful education reimbursement plan or something else — there are other things in this equation that aren’t just base pay, and I would talk about those things. Talk about equity, or options if there’s stock involved. Talk about those things in that context because the AI answer they’re getting generally is not going to account for those.

You could say, “Hey, we’ve actually looked at the data, and we’re in the top 30% of all companies in this market for this type of job. So, I want you to understand that we’ve actually done our due diligence here. We spent a lot of time and effort making sure we’re doing this the right way.”

If I was the employee, and someone told me that, I would feel much more confident, much more certain that the company has my best interests at heart and that they’re doing the right things when it comes to pay.

Editor’s Note: Additional Content

For more information and resources related to this article, see the pages below, which offer quick access to all WorldatWork content on these topics:

• Employee Compensation
• Finance and Budgeting
• People Analytics
• Total Rewards