AI capable of removing biases in wage negotiations: Pactum AI CEO

Negotiating a salary and benefits can be one of the most stressful parts of an employee's job. In the dawn of wide-scale AI adoption, employers are considering using artificial intelligence to work through the negotiation process more efficiently.

Pactum AI CEO Martin Rand joins Yahoo Finance Live to discuss how AI models can be programmed without certain biases, such as the gender pay gap when speaking with workers.

"When a new employee joins, they will get a chance to negotiate with an AI bot that negotiates kind of like a human, but we take all of the biases out of it," Rand says, citing statistics that women workers may engage less in negotiations. "Now, AI bots don't have this expectation. So, it removes that [gender wage] gap. And when our people go through the negotiations, 85% prefer to have this negotiation again with a bot in the future."

Rand outlines Pactum AI's opportunities to support other financial transactions and the company's ability to scale this product out for more employers.

Video transcript

[AUDIO LOGO]

SEANA SMITH: Negotiating a salary can be tricky to navigate. The stakes are usually very high. And it's a stressful situation for almost everyone involved. But what if AI can help? Pactum AI is doing just that. They're using generative technology to negotiate packages for their employees as well as big business vendor contracts, too.

Co-founder and CEO of Pactum, Martin Rand, joining us now. Martin, it's good to see you. This headline caught a lot of people's attention, because, I think, most people dread having those salary negotiations. There's a lot of stress. People worry about the right things to say. So you guys are taking a bit of a different approach by using AI. How does this work?

MARTIN RAND: Yes, indeed. And we might be the only company in the world doing this [AUDIO OUT] start thinking about that, because what we're doing basically is we're using AI to negotiate supplier deals for the world's largest enterprises, like Walmart and Maersk, unlocking hundreds of millions of dollars. So we have all of this technology. At some point, we thought, hold on, let's use it for our employment negotiations. And it's a chatbot negotiating with every new employee. And the outcomes have been pretty spectacular.

AKIKO FUJITA: So how does it work?

MARTIN RAND: So when a new employee joins, they will get a chance to negotiate with an AI bot. That negotiates kind of like a human. But we take all of the biases out of it. So if one of the major problems is a gender pay gap, pay gap, basically the problem is that women don't negotiate enough. Women negotiate only when they are expected to negotiate. And they negotiate a lot less than men, science has shown.

But now, AI bots don't have this expectation. So it removes that gap. And when our people go through the negotiations, 85% prefer to have this negotiation again with the bot in the future. And this is basically because every person can think what is the best outcome for them.

And if you think about it, employment negotiations are complicated, because there are so many things to talk about. There are payment intervals, health insurance, education stipends, titles, student loan support. And the list goes on and on and on. Essentially, there are millions of combinations for good deals. And a machine will find those deals much better than humans.

AKIKO FUJITA: Martin, I take your point about removing the bias in a human-to-human conversation. But as you know, there's also been criticism about biases built into AI and the need for more transparency on how it was actually programmed. So even if we're talking machine learning, how do you ensure there's no biases in that?

MARTIN RAND: So this is a little bit different. What the biases people usually talk about with regards to AI is biases that come from the training data. But this-- our system doesn't require a lot of such training data. Our system has been trained to find Pareto optimal outcomes in negotiations, meaning it'll know what is the best combination between tradeable terms. And there are no biases in these tradable terms per se.

SEANA SMITH: Martin, do you think that this is something that can be used at scale? Or are there some difficulties associated with this?

MARTIN RAND: It can definitely be used at scale. We're currently not offering this at a commercial level, simply because supplier negotiations are much more sought after currently. But in the future, definitely. And what our employees said when they went through these negotiations is that they are impartial, structured, efficient, and fast. So people will enjoy them a lot, and so will enterprises that will simply get a better deal as well.

AKIKO FUJITA: So it's a little more efficient. But I will ask what I think a lot of people are wondering is, isn't there a human component though that's necessary in these type of negotiations? I mean, you can remove bias. But also, isn't this part of-- I guess the question is, do we completely remove the human interaction out of it? Is that also a positive?

MARTIN RAND: We shouldn't completely remove the human interaction out of it. But-- but many people don't like negotiating. Many people are not professionals in negotiating, especially if they're negotiating something for themselves and especially when the concern is around money. So if we remove the human aspect only in that section, it'll benefit generally.

SEANA SMITH: Certainly intriguing. All right, Martin Rand, CEO and co-founder of Pactum AI, thanks.

MARTIN RAND: Thanks so much.