Employees are asking questions about working with AI tools at work, including who is accountable for mistakes. ADP’s Kasara Weinrich and Helena Almeida discuss how human judgment remains the key to using AI responsibly. For more, visit ADP.com/SPARK

Video Transcript:

Kasara Weinrich

Well, I'm excited to continue talking to you about obviously, we have organizations not just implementing ADP products, but they are really trying to determine how does an HCM product fit into our ecosystem alongside all of these other tools that we need to give to our teams and to our workers? And so we've gotten some questions from workers around the use of AI, kind of regardless of where where it sits.

And the first question, this is one- again, I'm so happy we have you here to to answer it. Not to know pressure I know, I know. Take a deep breath. Settle in. I'm just kidding. What protections do I have if AI makes a mistake that affects my work?

Helena Almeida
That is a good question. So first I would say our job, all of our jobs as workers who are using AI is to be the ultimate decider, right? It's up to us to make sure that we're looking at what the AI is putting out, figuring out if it makes sense, applying our own common sense, our expertise, our judgment to make sure that whatever the output is, is something that I want to use and I want to adopt. What I think might happen, right, is if somebody sees output from AI and they automatically think that it's accurate, right?

Because it looks very polished and it looks, you know, it has 16 bullet points and a, you know, a introduction, a conclusion. And people tend to have this. It's called automation bias. Right. The computer spit it out. So they want to rely on it. Our job is to make sure that we're actually questioning what the output is and making sure that it makes sense.

So if there is an issue with AI that impacts somebody's work, I think, you know, hopefully we can avoid that by having the worker kind of keep accountable, be accountable and keep in the loop and keep making sure that they're checking it.

But the flip side of that is what if something goes wrong in AI that impacts a worker negatively, for example, they feel like their performance review was assisted with AI and it was- it kind of didn't reflect their full contribution. That's definitely a time where I think workers should raise their hand, ask questions. The important thing here is AI is not the final answer.

And what AI needs to do is what what organizations need to do is really make sure that they are providing workers with information about how a AI is going to be used, how AI is going to be checked. But in the end, if a worker feels like they're being impacted negatively by something involving AI, I think it's it's on them to really raise their hand and make sure that they get to the bottom of whatever that issue is and clarify what their concerns are.

Kasara Weinrich
Yeah, I think that also ties back, and I love that you took that question two ways. So not just as a technology user. If I copy and paste an answer and send it off and the answer was wrong. There's one way that I would feel like, you know, am I going to be okay if I do that? And the answer is probably don't, don't do that.

Right. But then the second part of that is, how do I know if AI is making decisions about me in a system that I can't see or that I don't touch?

Helena Almeida
one of the things that I worry about is that, you know, the worker who who was worried about the AI making a mistake, the concern is that they don't use it. Right. And then they don't get the benefits of it. So I, I, you know, I think schools I think organizations should really lean into teaching people how to use it because it's not going away.

It's not sort of, you know, when is AI going to be introduced into our lives? AI is here, so we all have to learn how to use it and learn how to use it responsibly. And I would worry that somebody would be afraid that it would make a mistake. And I think people need to trust themselves, right? They need to trust themselves that they have the experience and the expertise that, you know, an organization hired them to bring and rely on that, to use AI to help and to make sure that they have the confidence that they're going to be able to to catch something.

The answer isn't not to use it. Right. It's just to use it smarter.

Kasara Weinrich
Yeah. And that actually, that makes a very good point from almost from a, from an education perspective, if you're just dipping your toe and if you're just kind of emerging into the space of this technology, kick it off with something you know really well. Yeah, right. Something that you can very easily fact check, something that isn't going to kind of make you pause and say, I don't know if this is true or not. Start with something, you know, and and kind of build the confidence and the technology in your own skill set.

Helena Almeida
That's a great that's a great piece of advice.