AI Hiring Under Fire: CHRO Kristen Duckett on Mobley v. Workday, Governance, and Accountability

Show Notes

Summary

What happens when an AI hiring tool is accused of discrimination—and your brand, hiring engine, and employee trust are all on the line?

In this episode, Rebecca Taylor sits down with Kristen Duckett, Chief Human Resources Officer at WorldStrides, to dissect the Mobley v. Workday case and what it means for HR leaders using AI in talent selection.

With a background in executive recruiting and hands-on experience leading an AI steering committee, Kristen details how to build real governance—not just policies on paper. She explains automation bias and why explainability features matter, how to audit outcomes without stifling innovation, and who to involve (CHRO, Legal, CIO, Product) at each step.

You’ll hear how to message internally and externally when the pressure is high, why HR must investigate before defending, and practical ways to pause high-risk AI features while keeping hiring moving.

Timestamps

[00:45] – New format: Real-life employee relations scenarios and why specifics matter

[02:21] – The Mobley v. Workday case: AI bias claims and employer liability

[03:02] – Governance first: CHRO accountability, training, access controls, and audits

[07:40] – Automation bias explained and using explainability features to spot issues

[10:32] – Building an AI steering committee: CHRO, Legal, CIO, and Product

[12:19] – When to involve recruiting: avoiding defensiveness while getting facts

[22:33] – Communicating under scrutiny: facts first, transparent updates, no platitudes

[28:36] – Avoid knee-jerk reactions: pause resume screening, keep low-risk AI (e.g., scheduling)

Takeaways

- Establish cross-functional AI governance—form a CHRO-led steering committee with Legal, CIO, and Product.

- Train recruiters to counter automation bias; use explainability features and require human judgment on final decisions.

- Audit outcomes regularly for disparate impact; don’t “set and forget” vendor tools—validate, test, and re-test.

- Communicate early and clearly: investigate before defending; share what you know, what you don’t, and what’s next.

- Pause high-risk features (like screening) while maintaining low-risk automations (like scheduling) to keep hiring moving.

- Own the outcome—CHROs are accountable for tool selection, safeguards, and continuous improvement, regardless of intent.

Sponsor

AllVoices brings all your employee relations work together in one place. No more jumping between spreadsheets, emails, and legacy systems just one place to document and manage reports, cases, investigations, and performance conversations. It helps you run a more consistent process, takes busywork off your plate with AI, and makes it easier to spot trends early, so you can work proactively, not just put out fires.

See a demo at ⁠⁠⁠⁠https://www.allvoices.co/

What is HR Voices?

HR Voices is a scenario-based podcast for People Leaders who’ve actually had to make the call.

Each episode brings experienced HR and People leaders into realistic, anonymized workplace scenarios—the kind you recognize immediately. Performance issues. Messy conflicts. Investigations that don’t fit neatly into a policy box. Instead of talking about their own companies, guests react to outside cases and walk through how they’d think it through in real time.

There are no right answers here. What you’ll hear is judgment: how seasoned leaders balance risk, fairness, legal reality, and humanity when the stakes are high and the path isn’t obvious.

HR Voices is for HR, People Ops, legal, and leaders who want to hear how other smart humans actually handle employee relations—without confidentiality breaches, hypotheticals that feel fake, or a lecture on “best practices.”

Rebecca Taylor (00:01)
Hello and welcome to HR Voices. My name's Rebecca Taylor. I'm your host and I'm here with Kristin Duckett. She's the CHRO at World Strides. And Kristin, thank you so much for being here, especially for our new and improved HR Voices format. Thank you for being our first guest to dive into this with us.

Kristen Duckett (00:18)
Thank you, I'm honored to be your first guest on this format.

Rebecca Taylor (00:21)
I think this is going to be fun because what we wanted to do was sort of we got some feedback that, you know, there's a lot of conversations that are happening that are very sort of top line thought leadershipy type conversations, which are good and valuable. ⁓ But our audience was really kind of looking for some more specifics around, you know, how to tackle certain employee relations type situations or even just HR scenarios that come up. So, you know, what we're doing is we're exploring real and fabricated anonymized employee relations scenarios through the lens of experienced HR and people leaders.

So we're going to be evaluating realistic workplace situations and demonstrate how we assess risk, apply judgment, and design practical responses. So the goal isn't necessarily for us to solve the world's problems here. We're really just talking about how really strong HR leaders think when we're facing ambiguity. And we're not necessarily looking for a single correct answer because a lot of times in HR, there just really isn't one correct answer. So ⁓ before we kind of dive into our talk track, ⁓

I think one of the things that is gonna be really important about this is that, you we're talking about real scenarios and also fabricated scenarios. So I wanna clarify that today's scenario that we're gonna be talking about is actually a real life scenario that's probably going to sound familiar to a lot of HR folks who are listening to this right now. ⁓ So what I'll do is I'll read the scenario and then Kristin and I are gonna kinda talk through if this came across our desk in the HR world, you know, how would we navigate it? What would we think about? And how would we kind of go about addressing it?

So, Kristen, you ready? All right. So the scenario we're talking through today is about discrimination in AI hiring, specifically Mobly versus Workday. So something that I know we in HR are tracking pretty closely because this is sort of an ongoing ⁓ case that's happening right now as we're even talking about it. So the story is a black job applicant applies for hundreds of positions using an AI-powered applicant tracking system and is rejected from all of them.

Kristen Duckett (01:51)
Great, I'm ready.

Rebecca Taylor (02:21)
He files a lawsuit alleging the AI screening tool discriminates on the basis of race, pointing to research showing that many AI hiring tools perpetuate historical biases embedded in training data. So the company argues the tool is neutral and bases decisions solely on qualifications. So this case raises questions about employer liability when AI tools, not human decision makers, screen out candidates from protected classes at disproportionate rates. So.

Before we jump to any solutions, before we start to talk into kind of the AI part, the hiring part, what stands out to you as most risky or most unclear in this situation?

Kristen Duckett (03:02)
I would say AI tools are not inherently discriminatory. However, it does not abdicate the user responsibility. And especially, it does not mean that they are ⁓ not responsible for the governance that really is, I think, part of using these tools. It's part and parcel. So I would want to know, first off, what was the governance process of these companies? I think oftentimes there's this...

rush to kind of jump into using AI tools. Everyone's hearing about it. Everyone's wondering how can I make my workforce more efficient? What can I be using to help my employees enjoy their work, be more effective? And so I do think there's this FOMO that we're seeing of I need to be using it constantly. And in that people rush to implement it, but don't really have the governance around it to support it. That's so important. And so, you know, as

leaders, not, it's part of our job to really make sure that we're kind of exploring the sandbox for our employees and making sure we're kind of putting those parameters in place so that they feel comfortable using these tools. And we don't, it's a fine line between, don't want to stifle innovation by having so much governance that it feels prohibitive, but you also don't want to let it just run wild. It's like giving a young child a bike for the first time without any kind of training wheels or any training. They might stay up for a little bit, but they're probably going to crash at some point.

and you don't want that to be your company.

Rebecca Taylor (04:32)
I think that's one of the things that makes this really hard too, is that we're all kind of trying to adopt new technology and new tools while we're still figuring out the governance, the capabilities of these tools, the risks of these tools. It's like at the end of the day, when there's fallout went from something like this, there's human lives that are impacted, right? It's not like someone makes a bad sale. It's like someone doesn't get a job that they might be able to actually be qualified for.

Kristen Duckett (04:58)
And I used to think, I cut my teeth in HR and executive recruiting. So I have a recruiting background. And I used to think before kind of in the last couple of years when these AI tools really were coming into favor that recruiting back then was much harder because we didn't have things like LinkedIn. We had binders of people we had to search through. Actually, I think recruiting can be harder today and just HR can be harder today in the sense that you have a lot more at risk and using these tools. And so you have to make sure that you're kind of slowing down to speed up.

⁓ that you are putting in place the right framework, the governance framework, making sure you're controlling access. Like who actually can use these tools? Do you have kind of a list of people at your company who are responsible for these tools? ⁓ Are you aware of the algorithm? you you use an external vendor and they have an AI, they have an algorithm for screening candidates. Do you understand how that works? Do you know if they do any bias testing? I mean, it's incumbent upon the leader to really be

asking these questions. And then do you have an audit process afterwards? Are you actually looking back and saying, okay, over the last six months, 12 months, let's look at the return. Like, do we find that there are, it's still showing a certain number of candidates from a certain population? You have to be looking into that and not just kind of setting and forgetting, which I think is easy to do when we're all so busy. We have so much on our plates. I think sometimes it can be this temptation of great, now I can bring in this tool, let it run. It'll make my workforce more

but then it can get you on the back end if you're not careful in how you set it up.

Rebecca Taylor (06:33)
Yeah, and there's so much data now that I don't know that sometimes it's like, you know, there's garbage in, garbage out, right? And with a lot of these AI tools, they can be really good at identifying patterns in data, but I don't think we as people yet are good at determining what data is relevant for certain things and what isn't. And that's, I think, kind of what makes some of this sort of complicated too. And so if you were in this scenario or close to this scenario,

⁓ You kind of started to get into this, I want to kind of dig a little bit deeper. So what would be some of the first things that you'd want to understand?

Kristen Duckett (07:10)
In terms of this scenario in particular, I would want to understand, like I said, the companies that were involved and implicated, did they have any kind of governance model set up? Did they have a steering committee that looks at what tools the company is using? Who did they have as the decision makers? Have they trained their recruiting team on using these tools? mean, these tools are meant to inform decisions. They are not meant to be the decision maker. And in fact, there's this phenomenon called automation bias.

Rebecca Taylor (07:12)
Yeah.

Kristen Duckett (07:40)
some researchers at the University of Washington discovered this, that people have a tendency to trust what AI algorithms will produce. ⁓ And they found that they did a study with a team of researchers and recruiters, and they were more inclined to trust the ranking given to them by the AI tool, even if they suspected that it might be flawed. And that to me is fascinating. And that's something we have to work against, because if human nature is such that there's this inherent trust in these tools, we have to

teaching our teams and our recruiters that you need to also exercise judgment, that it's not responsible for making the choice. You have to be able to do that. So I would want to know, like, what's the training regimen for these recruiters using these tools? Are managers checking in? You know, are they actually

understanding how the rankings are, if they're using it for candidate ranking, there's also ⁓ some of these tools have things called explainability features that you can turn on that actually shows you here's why the tool ranked candidates the way they did. And that gives recruiters an insight to say, okay, actually, that's not what I would have thought. Let me dig a little deeper. And that can uncover if there is any inherent bias in the model. But again, it's

What was their governance model? Did they actually have someone accountable? And ultimately, I would say the head of HR, the CHRO is if you're going to introduce these tools, you are the accountable one. It's not the company that you're using. Did they have any kind of ⁓ process by which they trained their recruiters? ⁓ Did they go back and audit their work later and say, this actually is what we would have wanted to see? Are they using ⁓

Rebecca Taylor (09:10)
Yeah.

Kristen Duckett (09:26)
the data that they're using, there's another example where there was a large tech company, we can all probably guess which one that is, about 10 years ago that was using an AI tool for recruiting and they had it trained on what would you think be logical past successful candidates. If we had the success with past candidates, we wanna use it. Well, it turns out most of these successful candidates happen to be men. And so it was not inherently biased, but it was because they set up the parameters to look at past candidates and not really thinking about

Rebecca Taylor (09:36)
Mm-hmm.

Kristen Duckett (09:56)
what does that pool look like? It started to disqualify resumes that had like women's chess club and women's engineering club. And again, it wasn't malicious or intentional, but it's what it was taught to do. And it will do what you taught it to do, yeah.

Rebecca Taylor (10:07)
It's what you taught it to do. Yeah.

Yeah. Yeah. And so it's kind of like, in your journey of trying to understand some of the governance around the tools, who do you need to hear from? who do you avoid hearing from too early? So when do you kind of start some of the actual conversations to start to understand this stuff and who are the people you're going to?

Kristen Duckett (10:32)
I you always I think it's always important for HR to have a strong relationship with with the organization with the CIO. ⁓

you know, anyone who's handling data on the company and think our ⁓ general counsel. In fact, at World Strides, we have an AI steering committee that is composed of myself, our general counsel and our assistant general counsel, our CIO and our chief product officer. And so we're kind of this four legged stool. And so I think it's important to make sure you have ⁓ others involved in the decision making process and have visibility into it, because they may see something that you just may not or you may not be seeing. So I think making sure you have

really strong relationship with your IT partner, your CIO is really key, general counsel legal. Who would I not want to involve too early?

I think it's often finance. We don't want to get finance involved. kidding. ⁓ No, I think it's important to honestly, I welcome people questioning kind of our methods and how we're using tools. I wouldn't want to say I'm not going to talk to you about this right now. ⁓ Let me do my thing and then you can see it later. I think you also want to be careful though.

with employees that maybe are newer to using AI tools, just making sure that because a lot of times there are assumptions made. They hear something in the news about AI bias or AI is discriminatory. And so there can be these assumptions made. So I'd much rather kind of lay the foundation and get the governance model set up first, be able to explain, no, we're we're being super intentional about how we use this tool. We're really checking into it. We're making sure that it's not the ultimate decision maker so that you can give

the entire company some assurance ⁓ that you're using the tools in the right way.

Rebecca Taylor (12:19)
Yeah, I think if it were me too, and I don't know if this is controversial or the right way to kind of think about it, but I would want to avoid talking to the recruiting team first because they're naturally going to come in defensive, right? Like I almost think that you need to kind of look objectively at how is the tool working? How is it set up and try to kind of gain an understanding before you involve people who may have been the ones to set it up or might,

be the ones that are maybe automating things incorrectly. I think you need to talk to the recruiters pretty early, but not first. They're like, the governing committee might be 1A and the recruiters might be 1B, right? Because naturally, they're going to come in defensive, as anybody would, right? ⁓ And they're going to have a valuable perspective. But I think we need to kind of understand

as objectively as possible, like what exactly is it that's happening? So we can even see, is the tool even doing this? Like, is this really happening? And what are the different pieces of it that can prove that whether it's happening or not?

Kristen Duckett (13:26)
And yeah, I think that's a great point. I mean, also the recruiting head, whoever's leading the team, I think that's someone that I would actually want to talk to pretty early on to understand, like I said before, the training and how are they engaging it. Because we don't want to put this on the recruiters. They're using the tool, but they're not necessarily, they should not be left holding the bag. It's really the leaders of the company who have decided to implement these tools. They're the ones that really have the responsibility to lead through this and to make sure that they're providing the right guidance and

Rebecca Taylor (13:31)
Mm-hmm.

Yeah.

Yeah.

Kristen Duckett (13:56)
So

absolutely, I would not want recruiters to feel that this is on them. But at the same time, the accountability does then trickle down, though. As I think we get more familiar with the capabilities of these tools, using them.

there does become some shared accountability then. Like if I have a recruiter who pinpoints something wrong or sees something wrong, I would want them to raise their hand. Now hopefully we create an environment where they feel comfortable raising their hand and saying, hey, I think this might not be working the way we intended to. But ultimately, again, leadership is responsible for creating that environment where people can come forward and raise their hand if they think something's wrong.

Rebecca Taylor (14:36)
Yeah, yeah. And like what assumptions would you have to be careful to not make at this stage? Like I know that we wanna, I'm gonna ⁓ say we wanna assume good intentions, right? Like we don't wanna assume that there's malice here, but are there other assumptions that you need to be careful of not making this early on in the process?

Kristen Duckett (14:54)
that it was the business pushing the recruiters to work faster and that's why they had to, they use these tools and they weren't as careful as they should have been. You know, I think sometimes in situations like this, the blame can kind of get, it's a hot potato, can kind of get shifted around. So I think it's the assumption that ultimately recruiting did have the responsibility. I think the assumption that,

perhaps the candidate was qualified, maybe the candidate actually wasn't qualified. I don't know if the candidate, I mean, there were a number of places where this person applied. The odds of this person not being qualified at all of them does seem unusual. But I think the assumption that there is any truth on either side before you start to dig in and understand the parameters of the, know, in terms of the governance and what the qualifications of the candidate were, I would want to make sure that before assuming anything, because I think you're right, there could be a defensive posture from the company to say,

⁓

person wasn't qualified anyway, so we would never have hired that person.

You may have actually hired that person if you had known, and so I think just making sure you come at it with a really clear eyed perspective ⁓ and that the goal is really to improve processes and make sure it doesn't happen again. And I think a lot of times in these cases we're all learning these tools now at the same time. What really matters is do you learn from these mistakes? Do you then take those those learnings and apply it so that you can evolve your processes, evolve your governance? That's really important.

Rebecca Taylor (16:22)
Yeah, I agree. in terms of mistakes, too, what do you think are some of the big mistakes that HR teams make in scenarios like this?

Kristen Duckett (16:32)
I think they rush into using these tools because...

Like I mentioned before, there's this fear that if you don't jump on the bandwagon fast, that you're going to be left behind or that you're going to be seen as not strategic or not tech forward. And so I think there's a mistake of trying to implement too fast without having, like I said, the proper governance or having kind of the right training. I think that's really important. I think right now 70 % of companies, I believe, are using these AI hiring tools. 99 % of the Fortune 500

use these tools. So it is widespread. And so we're going to see challenges like this pop up. ⁓ But I think it's really incumbent upon the leadership, as I said before, to just make sure you are thoughtful and intentional about how you use it. then you can start to really gain speed and gain traction and gain scale once you have the parameter set and once you have the foundation laid.

Rebecca Taylor (17:34)
Yeah, yeah. I also think in scenarios like this where there is a claim being made or, you know, there is, whether it's from a candidate or if it's from an employee, I think that there's a lot of HR scenarios where HR folks will assume that they have to defend the company first or that, you know, that like the company is right and this person is wrong because this, you know, the claimant is trying to take the company down. Like there, you know, there's so much discourse about

bad HR actors or HR isn't your friend, HR is here to defend the company. And I think that that is a mistake that some HR folks might make in this situation where instead of stepping back and seeing how real is this, is this, what is going on here? Is this an isolated incident? Is this something that's happening in other areas too? Like my first thought would be like, my God, this is the first person that says something, but is this the first person that's happened to, did this even happen?

And I think like there's a tendency to kind of just defend instead of investigate. And that's where a lot of these things tend to snowball too. And it kind of leads to that disenfranchisement that people feel with HR, to be honest.

Kristen Duckett (18:47)
Yeah, because usually I and other other places to like learning and development, right? In terms of career pathing. And so it would make you want to look and say. Okay, are there other places in the HR kind of realm where we're using these tools where there might be a bias. Inherent bias as well are people being cast over for promotion because the data is is working against them. So I think it would make you kind of.

Rebecca Taylor (19:02)
Yeah.

Kristen Duckett (19:10)
You know, it's overused phrase, but peel back the layer of the onion and just say like, okay, what else are we uncovering here? ⁓ But you're right there. They're 10. think there can be a reaction. We're all human beings. And if it feels like we did something that was inappropriate, or we didn't take our time, it can feel like a retribution against us. But that's how you learn. And that's how you that's how you build trust with your employees too. Because obviously, this is a very public is a very public lawsuit. And so I would want to be able to.

to

as the head of HR there, like explain.

Okay, here's what we're doing about it. We recognize we work. If it was found, I it's still ongoing, but if it was found that there were some, there is some probability there, acknowledge it and then talk about how you're moving forward. Because I am finding that there's some inherent distrust among some employees around AI or just this feeling of it's bad for the environment. ⁓ There's ethical issues, all of that. And so when you add things like this on top of it, it just creates this snowball of resistance. And that's not what you want. You want people to really understand

understand how these tools work, how they're being deployed, how they're being used, how they're being automated, all of that, so that there is this sense of, this is a tool that's meant to enhance your work, take away the repetitive things that you do all the time, give you free up your time if you're a recruiter to spend more time with candidates, building those relationships. Recruiters are often the first face of the company to a candidate, and that's the experience. so that's a huge responsibility and a huge privilege.

I'd love for my recruiters to have more time with the candidates and less time pouring over resumes. And so really making sure people understand that it's a tool that's meant to enhance their work, but again, not replace decision-making.

Rebecca Taylor (20:57)
Yeah, yeah. And there's a lot of competing pressures here too, right? So at some point, HR has to kind of start to make a call or start to sort of say, like this is how, this is what's going on here. This is what's not going on here. Here's who we need to continue to involve. Here's how we need to act. So when you look at something like this, I mean, this is obviously a very public lawsuit that's happening right now. Not every HR scenario is this sort of big thing like this, but this one has external pressure. Yeah, thankfully.

Kristen Duckett (21:22)
Thanks, Laine.

Rebecca Taylor (21:26)
This one has the external pressure ⁓ from the company brand perspective, right? It's got the internal sort of dynamic that there's between mistrust of AI, sometimes mistrust of HR. It's sort of the spotlight exactly where you don't want it. ⁓ And so, you know, when you're looking at something that you're kind of navigating all these competing pressures, ⁓ how do you weigh them? Like, what is, you know,

What sort of the, between the external pressure of the lawsuit, the internal trust that you're trying to rebuild with the company ⁓ and the hiring that you're still trying to do, how do you sort of weigh them?

Kristen Duckett (22:05)
I don't think we often have the luxury of kind of working sequentially and saying, okay, I'm going to tackle the external brand problem first, and then I'm going to tackle internally how my employees are feeling. So that's, think, in many ⁓ leadership roles, you have to kind of juggle all these things at once. And so I think the first thing is to really, we talked about this earlier, dig in and say, like, was there an issue? Did we actually miss the mark here? ⁓ Did we actually just roll out this tool and we could have had

Rebecca Taylor (22:15)
Wouldn't that be nice?

Kristen Duckett (22:33)
better parameters in place, we could have really made sure we trained her. And so really understanding the situation. I think that's the most, what are you working with? Because you can't start to spin something or.

build a narrative around something if you actually don't know what you're talking about, because then you completely lose trust. If you're trying to build this narrative and trying to tread water and then you found out that you actually were ⁓ culpable of something, you completely lose credibility. think there is also this sense of like people are willing to give others a second chance if they can acknowledge that they

Or at fault if they were fault ⁓ and then talk about what they can do. So I think it's there is that pressure though. I think sometimes, especially for publicly traded companies ⁓ to respond quickly and to get the talking points out or to get the memos out of like, ⁓ you know, here's what we're doing or we didn't do anything wrong and here's why. ⁓ But I think I would caution companies to really make sure that you understand what you're dealing with first, because if it comes out later that there was a problem again, the credibility is lost ⁓ at the same time.

You know, have to also have messaging to your employees and make sure they understand that this is something we're working through. are trying to understand, you know, the root cause, what happened, and commit to the employees that we are going to do something different going forward, if in fact there was an issue.

Rebecca Taylor (23:53)
Yeah, I'm with you. It's like the messaging that you put out there is the thing that could make or break internally and externally any sense of trust. Because if you're putting out sort of platitudes that are just like, well, you we operate with integrity and we do this. It's like, OK, everyone's going to read right through that. This is kind of where, especially internally, you know, I know that there are a lot of nuances in how you communicate different things at different levels of the organization and transparency and all of that, right? Especially at an organization the size of work.

But I do think being transparent as much as you can with your employees saying, hey, this may or may not be a bigger issue. This may be something that this is something that we're currently tackling and investigating and figuring out. We'll keep you posted as we go. don't know everything. Just even saying that quickly is, think, something that can kind of go a long way where employees might say, OK, this situation is unfortunate. ⁓

but I trust that the people who are in charge of fixing it are at least gonna go about it in the best way possible.

Kristen Duckett (24:56)
All right, like we take this seriously, like giving assurances that we're not just brushing this under the rug, that we are serious about uncovering what happened. And to your point, we may not know exactly at that moment, but silence is always often the worst approach because people fill in the gaps, they create their own narratives. So you want to make sure you get in front of that first, get in front of the issue really, really quickly. Let employees, let investors, whomever know that you are on top of it. You are looking into it. You are committed to taking action.

Rebecca Taylor (24:59)
Yeah.

Kristen Duckett (25:27)
And that gives I think people assurances that this situation is meriting the attention that it really deserves.

Rebecca Taylor (25:34)
Yeah, yeah. And I think we kind of started to touch on this, but just for the sake of coming to the point where you're doing your investigation, you're figuring out all the information, you're talking to all the people. At the end of the day, who is the one who owns the outcome? Like, let's say it turns out, I know that this isn't a binary scenario, right? But like, let's say there's scenario A, there's outcome A where there was no foul play and it was just a glitch. And there's scenario B where

there was the tool was biased and it was making decisions and it was implemented improperly. Either scenario or either outcome, who owns that outcome?

Kristen Duckett (26:12)
I think ultimately it's the head of HR. If they're the person bringing in the tools and signing off on the budget, including the tools in the budget and making the choice to ⁓ engage with that third party or use the tool internally, ⁓ I think ultimately you can't be a leader if you're going to push the blame onto someone else. ⁓ And so I think that's why there is a responsibility that we have to, again, not stifle innovation because you don't want me, you want to this feeling of fear where if I make a mistake,

I'm going to get fired or if I do that, because I think that's dangerous territory too. But ultimately, whether it's found that the person, and especially if it was found that there was just careless kind of ⁓ treatment of the tool where, we'll figure it out later, just go ahead and use it, it carries different weight. think if the leader really was trying and trying and didn't know and thought they were doing the right thing and then came to find out that they weren't, but there is a remorse there.

Rebecca Taylor (27:00)
Yeah.

Kristen Duckett (27:12)
and there's a willingness to improve the processes and learn from that. think it's viewed very differently. But ultimately, to answer your question, think as heads of HR, we absolutely are responsible for anything that could happen within the realm of our organization.

Rebecca Taylor (27:29)
Yeah.

And it's like intent versus impact. It's like, even if you didn't mean for things to go poorly, the impact is that if they, if, if it did, then you are responsible for that impact, even if the intention wasn't there and levels of responsibility that can mean so many different things, right? Like, you know, does someone get fired? Does someone have to like pay up? You know, what's the, there's, there's too many sort of, there, there's too many different ways that that accountability can sort of live. But, you know, I think that some of the

Kristen Duckett (27:34)
Yeah.

Right.

Rebecca Taylor (28:00)
Some of the patterns that you see in situations like this too, though, is sort of almost over-correcting, at least I think. So if you're in a situation where tools that you've implemented are under fire because for this reason, ⁓ it can be very tempting to sort of over-correct and say, OK, well, we're just going to pull AI out of everything. Like, we're just going to stop it completely until we figure this out. Now, lawsuits like this can go on for years, right? So what are some other patterns that we kind of see?

companies falling into if there is something that's ongoing, but they don't have the luxury of just hitting stop.

Kristen Duckett (28:36)
I mean, recruiting has to keep going, right? And I think also if you've introduced these tools to recruiters and they've gotten used to adjusting their ways of working around them, you don't just want to rip that out and say, now we're going to see. Because then there's also this lack of trust around, well, do I really want to get used to using something only to have it get taken away if it's found that it's not working well? So I think it's, yeah, the knee jerk reaction doesn't go over well either. And I think it's just, you know, it might be like, hey, we're going to stop using it to screen out resumes.

to use keep using it to schedule candidates. There's elements that you can still continue that are fairly innocuous. And that's what I would say would be important is like, how do you keep it? How do you keep the train running? Maybe slow the train down a little bit, but you keep the train running, because it can't stop once you stop to get it going again, takes a lot of work. And you have to rebuild that momentum, rebuild that trust. And so I don't think the answer is to just pull it completely. Although I can understand why there might be that reaction, especially if it's as glossy like this, but better off to kind of just

evaluate where you can keep it working in a much more low risk area.

Rebecca Taylor (29:41)
Yeah, I think that kind of segues into sort of some closing thoughts for this conversation too is what advice would you have for an HR leader who's in maybe not this exact scenario, but in a scenario like this where your AI tool is maybe potentially doing something it isn't supposed to, what's advice that you have for HR leaders who are kind of in this scenario?

Kristen Duckett (30:08)
in the decision-making pool or using the tool currently, like get a grasp on like where this tool is being used and how it's being used and then understanding was there intentional kind of, you was there just misinformation or just lack of training? What was the kind of the source of the problem? And then maybe you pause that. So again, if it's something around, it's screening, it has bias in screening, pause that right

now and let yourself kind of figure out where to to make improvements. ⁓ And I would also say like it's important to look at how this how AA is just really evaluating or helping us evaluate the ways of working and how we how we work today. It's not meant to just be a know plot you don't plop it on top of existing processes. I would really encourage teams to say like how do we want to rethink how we work and how work gets done and then use the tool to amplify that and bolster that. I think too often it's just you know people

throw it into the mix because everyone's so busy and then it's just left to run wild. So I think I would encourage leaders to really just take a beat, ⁓ be really clear with their leadership team, with their peers, what's going on and what they're doing, that they're investigating, that they're looking into what the problem might be, that they're committed to enhancing processes, enhancing training, all of that. And just giving the sense of ownership, like I've got this, this might be a problem but I've got this, I'm committed to fixing it, I'm committed to

to keeping us using AI because that is the future. We want to make sure that we do it, but we do it in the right way.

Rebecca Taylor (31:45)
Yeah, I love that. It's like, a beat to sort of like respond, don't react, right? It's like, you know, don't immediately just sort of get defensive about it. Make sure that you're doing the things that kind of help you understand what it is that's going on. And I like the point about, you know, take responsibility in an empowerment kind of way. It's just like, I got this. I'm going to figure this out. I'm going to bring in the people that I need. You know, don't go into the whole like deflection blame game. It's never that that's never.

that's never done well for people.

Kristen Duckett (32:17)
I tell my kids and my husband always says this to the cover up is worse than the crime. So come forward early. Let us help you work, work through it. But if you find out later that you covered it up, that's never, that never works out well.

Rebecca Taylor (32:22)
Yeah. Yep.

Yeah, that's like the part. I'd love that. We're going to end on that. ⁓ Thank you, Kristin, for talking us through this. know, you know, I'm going to be following this like pretty closely as I think anybody who's kind of listening to this is because I'm sure there's going to be some precedent set with whatever outcome is of this suit. ⁓ And so, you know, keep your eyes on what's going on and thank you for listening to this episode. Have a good day, everybody.

Kristen Duckett (32:31)
you