The New AI Pay ‘Experts’

The New AI Pay ‘Experts’

NOTE: While this transcript has been reviewed, it may contain errors. Please review the episode audio before quoting from this transcript.

Jill Wiltfong:

Hi, I’m Jill Wiltfong, Chief Marketing Officer for Korn Ferry, and this is Briefings, our deep dive into topics that corporate leaders need to care about.

Say you’re about to ask for a raise or negotiate a new job offer. Wouldn’t it be great to know exactly what the firm can really offer if you push?

That’s what more workers are hoping for in today’s wild AI era. New data finds that workers are sending, get this, three million messages a day to ChatGPT about wages and compensation. And it doesn’t stop there. AI is helping them prepare exactly what to say.

This AI upgrade arrives as firms move away from “peanut butter raises”—spreading pay increases evenly, but thinly, across everyone—and toward more competitive, Hunger Games-style battles for a bigger share.

But could this be a science fiction story gone wrong? It turns out AI’s wage data can be off, and it can’t know what your company’s compensation plan looks like in any given year.

So before you go prancing into your boss’s office, let’s take a closer look at the pros and cons of these new AI pay experts.

Before we start, if you’re watching us on YouTube, please be sure to like, subscribe, and leave a comment to let us know your thoughts on this topic.

I’m joined now by Tom McMullen, Korn Ferry’s North America leader in its Total Rewards practice. He’s helped many company leaders navigate this new world of AI-fueled pay discussions. Tom, it’s really great to have you with me today.

Tom McMullen:

Jill, pleased to be here.

Jill Wiltfong:

So Tom, let’s get to the heart of the issue. You’ve said that the mistake some employees make when using AI to prep for salary discussions is that they view the data as doctrine, when in fact, it’s not. Elaborate on that for me.

Tom McMullen:

The market data someone can pull from AI is an input. It’s not the final verdict.

AI-generated compensation data reflects very broad market signals. It does not reflect a company’s specific job design, pay philosophy, financial constraints, or even how the organization sees an applicant or employee relative to the requirements of the role.

So from an employee perspective, it’s something to have in hand. But it has to be viewed as a data point, not the end-all, be-all.

Jill Wiltfong:

Okay. So salary data aside, you’ve also said you see a real benefit in how AI can help structure a pay conversation. I have to say, it’s a little ironic to think that a robot might be needed to help with this very human interaction.

So talk about where you see the advantage in these AI-scripted discussions. Where would you lean into those?

Tom McMullen:

AI can give you the number, but it can also give you a whole lot more.

It can offer recommendations on how to approach your boss in a tactful, professional way. I think it’s upping the game in terms of the conversations that will happen between management and employees.

Jill Wiltfong:

Let’s put AI to one side for a moment. You’ve mentioned there’s a generational shift in how workers approach pay today. We have, I think, five generations in the workforce right now, but you’ve said the youngest generation of workers, Gen Z, has a completely different view of pay conversations.

What do you mean by that? What are you seeing play out?

Tom McMullen:

For Gen Z, pay is a data point. It’s a reference point. It’s a piece of information they share freely with friends.

That’s very different from, say, baby boomers like me. For us, pay was confidential. It was secretive. You didn’t tell anybody about your pay.

Gen Z grew up with the internet. They grew up with social media, LinkedIn, Glassdoor, and all of these websites where the whole employee experience has opened up.

If I’m interested in Company ABC, I can see what other people think about working there, including pay, culture, and many other aspects of the organization.

Jill Wiltfong:

That’s a classic scene from the movie Jerry Maguire, where Cuba Gooding Jr. tells Tom Cruise to “show me the money.” It feels like a lot of these AI-derived pay conversations might be similarly direct.

But Tom, you’ve said that organizations often have not adequately equipped their managers and HR staff to respond to this new approach to negotiation. So what do you advise leaders to do right now to get their managers prepared for what is probably inevitable: these AI-powered, very open discussions?

Tom McMullen:

I think job number one is getting very clear with managers on how the pay system works.

One of the reasons some organizations don’t do this is because they’re not clear on the rules of the road or the frameworks themselves.

Who do we target our pay against? Are we pay leaders? Do we match the market? Do we lag the market? How do promotion guidelines work? Why are some jobs bigger or smaller than others?

If you want to have pay transparency, you have to start with the management group. You have to educate them and help them understand their role in these conversations.

Jill Wiltfong:

If a manager finds themselves confronted with surprise questions for which there are no immediate answers, what’s the best way they can respond without becoming defensive or sounding dismissive?

Tom McMullen:

It’s okay for managers not to have an answer in the moment. What matters is how they respond to the employee.

Part of that is acknowledging that you’ve heard them. Don’t argue. Acknowledge the question without necessarily committing to an answer right away. And if you don’t know the answer, have a process for getting back to that employee.

Commit to clear follow-up and next steps. A manager who stays curious and calm, rather than feeling guilty or defensive because they don’t have the perfect answer, is going to be much more effective.

Jill Wiltfong:

I want to end on something we mentioned at the top, which is that “peanut butter raise” theory that may actually be on the decline this year.

It’s never an easy thing for a workforce to hear that raises won’t be given out across the board. How should company leaders communicate this kind of message in a way that doesn’t deflate the whole team?

Tom McMullen:

When budgets tighten, I think the most responsible approach is to take those precious dollars and focus them where they matter most.

It’s also about being transparent in the messaging. I might not like the percentage I get, or the organization’s overall budget, but if I understand that times are challenging, how our company compares to others, and why the organization is making those decisions, that matters.

If the company is straight with me and treats me like an adult, I can handle that. I’d much rather hear that than hear very little, or be left with innuendo.

Jill Wiltfong:

All right, Tom, thank you so much for coming on and sharing your wealth of pay wisdom—pun intended. Appreciate it.

Tom McMullen:

Thank you.

Jill Wiltfong:

We’ve talked about how workers are increasingly using AI to prep for salary conversations. After the break, we’ll touch on a related question: Should AI determine employee pay?

Many employees say yes. More on that when we return. Stay with us.

Jill Wiltfong:

We’re back. In the first half of this episode, we talked about how more and more employees are turning to AI for salary negotiation help.

Now we turn to a related issue: Two-thirds of job candidates favor firms that use AI in pay decisions.

Here with me to dissect this angle is Mark Esposito, a professor of strategy and technology policy at Northeastern University. He’s lectured extensively on the role AI might play in corporations, so it’s great to get his take here.

Mark, wonderful to have you on.

Mark Esposito:

Absolutely, and good to see you again, Jill.

Jill Wiltfong:

That last clip featured hedge fund manager Ray Dalio encouraging people to allow AI to make decisions for them.

Many workers might agree. Not only do most candidates prefer firms that use AI in pay decisions, but one out of three employees trust AI over their managers to decide compensation.

It’s a striking stat, but Mark, you have a nuanced interpretation of that data. You say it more likely reflects a deep skepticism toward human managers rather than a love of AI. And you’ve said that skepticism is justified. What makes you say that?

Mark Esposito:

I don’t think this reflects a preference for AI over humans, or a belief that AI is inherently better. I think it reflects broken trust in management itself.

If you think about how managers have historically used compensation as a tool of control, and how blurred the boundaries can be around pay decisions, it makes sense that some employees might want to trust algorithms that feel less political in that specific process.

Jill Wiltfong:

There are certainly some upsides to using AI to determine pay, and you touched on a few of them there. Of course, there are also downsides.

Take me through the upsides first, and then let’s talk about the potential downsides.

Mark Esposito:

Transparency is a real upside, especially when it comes to AI.

If you have a system that is clearly calibrated around job-relevant factors—skill levels, roles, the complexity of the role, market rate, and tenure within the organization—you can quantify those factors in a more structured way.

That can help eliminate some of the hidden negotiations that often create resentment.

Jill Wiltfong:

So what do leaders need to do to watch out for some of the potential pitfalls of AI-based pay decision-making?

Mark Esposito:

The downside is that if the algorithm inherits bias from the data it has learned from, it can amplify that bias.

For example, if there has been historical pay discrimination based on gender, the model can simply reinforce that unless there is clear intervention to correct it.

Jill Wiltfong:

That’s a clip from Star Wars, where Luke Skywalker gets to know his soon-to-be invaluable robot companions.

So Mark, we’ve said there are upsides and downsides to letting AI have control over pay decisions. But is there a middle ground, where humans and AI work together like Luke and his droids? And if so, what does that compromise look like?

Mark Esposito:

I think the right balance is making sure we don’t create a pay system that is so highly structured and consistent that it starts creating a level of convergence that kills competition.

We need digital infrastructure, and we need clear mandates for what people are now calling algorithmic agency. But we also need to preserve a role for human judgment.

Humans should be able to audit decisions and make calls, even when they are counterintuitive. The goal is to build some degree of uniformity while also preserving the distinctiveness that humans bring to the conversation.

Jill Wiltfong:

Mark, always a pleasure. Thank you for sharing your thoughts here today.

Mark Esposito:

Thanks so much, Jill.

Jill Wiltfong

The executive producer of Briefings is Jonathan Dahl. Today’s episode was produced by Rupak Bhattacharyya and Zachary Dore, and it was edited by Jaren Henry McRae.

It contains reporting by Russell Pearlman, Ariane Cohen, Peter Lauria, and Meghan Walsh. Our video segment contains original artwork by Fraser Milton, Haley Kennel, Jonathan Pink, and Sasha Kotzek. Our web operations are managed by Ed McLaurin.

Don’t forget to read our magazine—available at newsstands and at kornferry.com/briefings.

That’s it for Korn Ferry Briefings. I’m Jill Wiltfong. See you next time.