Wellness
Rewards
Recognitions
Learn more about Wellable's all-in-one
employee engagement platform
AI Lawsuits Capturing HR’s Attention
Estimated reading time: 7 minutes
In this week’s episode, Nick and Geoff dig into two landmark AI employment lawsuits every HR leader should be following, and a new workforce sentiment story. Meta employees are suing over an AI-driven layoff process they claim targeted workers on protected leave. The Workday class action advances on the question of whether AI vendors or their employer clients bear liability for hiring bias. Lastly, a survey shows younger workers increasingly describe their relationship with their employer as a “situationship” with no long-term commitment on either side.
Pressed for time? Here’s a quick summary…
- Twenty-six current and former Meta employees allege the company used AI to score, rank, and select employees for termination without neutralizing protected leave as an input
- Meta’s spokesperson categorically denies AI made those decisions, creating a stark factual divide at the center of the case
- Mobley v. Workday is advancing through the courts on the argument that AI hiring tools that replace functions normally performed by employers carry commensurate liability
- Workday’s defense—that its tools only screen candidates and final decisions rest with employers—may be valid but puts its own customers on notice that they bear the liability for how AI outputs are used
- HR and IT leaders should be asking vendors directly how their tools handle protected class attributes, what bias testing has been conducted, and where liability sits before using AI in any consequential hiring or termination process
- 59% of workers report no clear long-term path at their current employer, and the resulting “situationship” is partly a job market story
- Employees cannot afford to leave but see no investment in their development either, a dynamic compounded by the hollowing out of middle management and the redirection of organizational energy toward AI over people
Episode Summary
The Meta Lawsuit: Did AI Decide Who Got Fired?
Meta has conducted some of the largest workforce reductions of the past two years, letting go of tens of thousands of employees across multiple rounds of cuts. A lawsuit filed by 26 current and former employees now alleges that those decisions were not made by managers with knowledge of their teams, but by a constellation of internal AI systems that scored, ranked, and selected employees for termination, and that those systems disproportionately targeted workers who had taken protected leave in the preceding two years.
The legal concept at the center of the claim is what the lawsuit describes as Meta’s failure to “neutralize the inputs.” In plain terms: if an AI system is evaluating employee performance and productivity, and an employee spent part of that period on medical leave, parental leave, or other protected leave, the system should account for that gap before drawing conclusions about that person’s output. The allegation is that Meta’s system did not.
Meta’s response is categorical: a company spokesperson stated that workforce management and organizational decisions were made by people, not AI. That framing creates a very clean factual divide. This is not a case where both sides are arguing about whether the right precautions were taken. One side is claiming AI drove the decisions. The other side is denying AI was used in a decision-making capacity. The court will ultimately have to determine which account is accurate, but the gap between the two positions is unusually wide.
Given Meta’s very public and ongoing commitment to AI across every aspect of the business, the claim that a major workforce restructuring was conducted entirely through human judgment, without any AI tooling in the process, will require convincing evidence to hold up.
The Workday Lawsuit: When the Vendor Becomes the Defendant
The Workday case, formally Mobley v. Workday, has been working through the courts and received renewed attention through reporting in Reuters and HR Brew. Where the Meta case is about how an internal AI system was used, the Workday case asks a different question: when an employer uses a third-party AI platform that produces biased outcomes in hiring, who bears the legal liability?
Workday’s defense is similar to Meta’s: its tools screen candidates, they do not make hiring decisions. Final decisions remain with the employer. Therefore, Workday argues, any downstream legal violation sits with the company that used the tool, not the company that built it.
The court has refused to dismiss the case, and the reasoning matters. The judge’s position, as Nick describes it, is that an AI platform that effectively replaces functions a company would otherwise perform internally carries some of the responsibility that would have attached to those functions. The vendor cannot fully insulate itself behind the argument that the tool only assists rather than decides, if the practical effect of that assistance is that consequential choices are being driven by algorithmic outputs.
Nick notes that there is something strategically odd about Workday’s defense. If the argument to employers is that they bear full liability for how the tool’s outputs are used in decisions, that is essentially a caveat emptor message to the customers Workday is trying to sell to. It may be legally defensible, but it is not a comfortable position for a company whose growth depends on employer trust.
The broader implication for HR leaders is practical. Both lawsuits expose a gap in how organizations are thinking about governance when AI tools are in the hiring or termination loop. Nick and Geoff both argue that HR leaders, alongside IT, need to be asking vendors direct questions: How does the tool handle protected class attributes? What bias testing has been conducted? Who bears responsibility if the tool’s outputs result in a disparate impact claim? Using AI in consequential employment decisions without clear answers to those questions is not a black box that insulates the organization from risk. It may be the opposite.
The Work Situationship: What Happens When Companies Stop Investing in People
The third story connects directly to the first two. A survey from Zety found that 59% of workers report having no clear long-term path and not looking forward to coming to work. Thirty-two percent say they have limited growth opportunities. Twenty-three percent describe feeling unfulfilled or frustrated. The term the research uses to describe the resulting dynamic is a “situationship,” borrowing from Gen Z relationship vocabulary: a casual arrangement with no clear commitment on either side
Geoff and Nick are both clear that this is not purely a generational attitude story. The situationship exists for structural reasons on both sides of the relationship. Employees cannot leave because the current job market is difficult enough that testing it feels genuinely risky. Many have heard stories of hundreds of applications with no responses. But they also do not see a long-term future at their current company, because the signals of investment in their growth, development conversations, mentorship, career path clarity, are not there.
Nick connects this directly to the middle management reduction trend. The managers who would traditionally have been responsible for those development conversations, who knew a young employee’s work well enough to sponsor them for a stretch assignment or advocate for their promotion, are the same layer being cut across the enterprise. And the energy that might otherwise go into talent development is being redirected toward AI agents and infrastructure. The math works in the short term. The long-term cost shows up years later, when companies look for experienced internal candidates to fill senior roles and find a bench that was never developed.
Frequently Asked Questions
Twenty-six current and former Meta employees have filed a lawsuit alleging that the company used a constellation of internal AI systems to score, rank, and select employees for termination during recent workforce reductions. The central claim is that these systems disproportionately affected workers who had taken protected leave, because the AI did not account for those absences when evaluating performance and productivity. Meta has denied that AI was used in a decision-making capacity, stating that workforce decisions were made by people.
The term comes directly from the Meta lawsuit. It refers to the practice of adjusting or excluding certain data points from an AI evaluation to prevent protected attributes from influencing the outcome. In this case, the claim is that Meta’s AI systems scored employees on productivity and performance without neutralizing the impact of protected leave, meaning an employee who was on medical or parental leave during the evaluation period would be scored lower through no fault of their own.
Mobley v. Workday is a class action lawsuit arguing that Workday’s AI-powered hiring tools produce biased outcomes that disproportionately affect certain protected groups. The case’s central legal question is whether Workday, as the vendor whose platform functions as a substitute for decisions that would otherwise be made by employers directly, carries some of the liability for those outcomes. The court has refused to dismiss the case, finding the argument that the tool’s role goes beyond mere assistance to be legally credible.
This is precisely what the Workday case is working to establish. Workday’s position is that its tools only screen candidates and that final decisions rest with employers, making employers the appropriate defendant. The court has pushed back on that framing, suggesting that a platform that effectively performs functions previously carried out by companies may carry commensurate responsibility. HR leaders should not assume that using a third-party AI tool insulates them from liability; the opposite may be true.
The term describes a casual, uncommitted relationship between employees and their employers, one where neither side is fully invested in the long term. A Zeti survey found that 59% of workers have no clear long-term path at their current employer. The situationship persists because employees feel unable to leave in a difficult job market but see no meaningful investment in their development either. Nick and Geoff connect it to the hollowing out of middle management and the redirection of organizational energy toward AI infrastructure rather than people development.
Nick and Geoff identify three practical steps. First, ask vendors directly how their tools handle protected class attributes and what bias testing has been conducted. Second, ensure that consequential employment decisions, both in hiring and in terminations, have a documented human review process that is not purely derivative of AI outputs. Third, work with IT and legal to assess whether current AI tool usage creates disparate impact exposure, particularly in any process that evaluates large numbers of employees simultaneously.
Full Episode Transcript
Nick: Welcome to the Wellable Weekly Podcast, where we talk about key topics and trends at the intersection of wellbeing, technology, and HR. Geoff, I can’t believe it’s the end of July.
Geoff: No, summer is flying by. It always happens like this though, right? We get into June, it starts to warm up, and then all of a sudden we’re wrapping up the end of summer.
Nick: I always feel like that’s the case, but this summer for some reason feels just a little bit quicker. We’re from Boston, so the end of summer means the coming of fall, which is very nice, but then also means the coming of winter. Lots of exciting stuff happening this summer, including maybe some not so exciting news if you’re part of this process — a couple of big lawsuits related to AI that are worth talking about because it’s really HR’s use of AI that’s generated these two major lawsuits against two big companies. We’re talking about Meta and Workday.
Meta is interesting because the lawsuit was filed by a group of current and former Meta employees — 26 of them — saying that the use of AI in these recent layoffs disproportionately affected certain workers, specifically workers who took protected leave in the last two years.
Geoff: If Meta really did miss this, how did they not consider that as part of the criteria when they ran their process to shortlist workers for potential termination? I know it sounds like they were using internal AI tools to make the process more efficient, and probably in their minds more accurate — the idea being that AI can produce a truly objective output without human bias. Meta has categorically denied any wrongdoing. But it seems like one of those things where if they did run an AI-led termination program, how could you not consider gaps in productivity due to medical leave? It just seems too big to miss.
Nick: It was someone’s job to think about something like this. The term in the lawsuit is that the company “did not neutralize the inputs.” To your point, protected leave would be an input you would need to consider when thinking about productivity over a period of time. What’s really interesting is that Meta’s spokesperson isn’t saying they used AI but took proper precautions. The claim says Meta used a constellation of internal AI systems to score, rank, and select employees for inclusion on the termination list. And the Meta spokesperson says: workforce management and organizational decisions were made by people, not AI. So one of these groups is simply right and the other is simply wrong. It’s a very clear factual divide.
Geoff: And it’s a very deliberate use of language. You can picture how they would use an internal AI system for a process like this, given their very public commitment to applying AI across all forms of their business. It’s not beyond reasonable doubt that they would try to do something similar for an internal process like this. And Meta is not the only organization to come under fire recently for the use of AI tools in people-related activities. Workday, which we touched on in the Stanford study episode, has a class action lawsuit continuing to move through the courts. A recent Reuters and HR Brew article unpacked the Mobley v. Workday case, and it’s really interesting to learn not only about the potential algorithmic biases at issue, but just how extensively companies are using AI in their talent acquisition processes today. That stat we mentioned before, that 90% of all companies are using AI in their talent acquisition process, keeps coming back up.
Nick: Workday is a very well-known HR platform. Many of our listeners are probably using it. And as with most of these AI tools, they make the process more efficient, but can introduce bias in a variety of ways. In the Workday case, the judge has refused to dismiss the lawsuit. The core argument the court found credible is that Workday’s AI function effectively replaces jobs and responsibilities that would normally be carried out by the company itself. By doing that, Workday takes on some of that responsibility.
Workday’s response is that their AI tools for screening resumes and applicants do not make hiring decisions. They simply help screen candidates. The ultimate decision falls on the employer. So they are arguing they’re not liable — the company using their tool is the one liable. Which is interesting, because you’re effectively telling your customers: use our product, and if there’s an issue, you should be the one being sued. That’s an odd argument to make to your own customer base.
The laws being touched here are significant: ADA, the Family and Medical Leave Act in Meta’s case, the Pregnancy Discrimination Act. AI is touching all of these because of the sheer volume of decisions being processed. The scale is what creates the legal exposure. I think HR leaders, alongside IT, need to be asking vendors direct questions about how their tools handle protected attributes, what bias testing has been done, and where liability sits if something goes wrong.
Geoff: All of that time and energy being poured into analyzing AI’s impact, both the efficiencies and the risks, means less focus on people. Good old-fashioned career development, mentoring, the tried-and-true methods of workforce management, are getting less attention. That connects to the last article I wanted to talk about: how work is feeling like a situationship. A casual dating scenario — not something in the vocabulary of two happily married guys, but it’s certainly how Gen Z and others are describing how they feel about their employer. A lack of commitment in both directions, where employees don’t feel the company has a long-term plan for them and they don’t feel loyalty or long-term commitment to the company either.
Nick: To put some color on it, the survey is from a company called Zeti. Some stats: 59% of workers say they have no clear long-term path and don’t look forward to their current workplace. 32% say they have limited growth opportunities. 23% report feeling unfulfilled or frustrated. That’s the situationship. They’re kind of just in a situation with their company. And the reason they stay is that they can’t leave. The job market isn’t super attractive. They’re hearing stories about people submitting a hundred applications and not getting one response. Some of it’s true, some is perception, but the perception is very clear: you don’t want to test the market right now. So they’re staying somewhere they don’t see a long-term future, but they’re not out there searching for what’s next.
We talked last week about middle managers being cut, but it’s the same concept. So many companies are focusing their best talent on developing AI agents rather than developing their people. You’re cutting the middle managers who could supplement talent development. You don’t feel the cost of that immediately, but eventually you come to your bench and ask: who’s our next up-and-coming star? Do they have the skills developed over the last five years to rise up? If you don’t invest now, you will feel it later. And the argument that you’ll just hire that talent externally misses something important: when you work with a young person for five or ten years, promote them to something senior, you have a history of knowledge about how they work and they have a history of knowing the company. That institutional alignment goes away when you only hire from outside.
Geoff: There was a quote in that article from an anonymous Workday executive, of all companies: AI may be rewriting the rules of work, but it cannot replace the value of engaged, motivated people. That sums it up well. When things are changing this quickly and the focus on technology is this strong, it’s harder and harder to keep people feeling genuinely invested. That seems like a good place to wrap up. Thanks as always to those who tune in. You can get Wellable Weekly on Apple Podcasts, Spotify, or wherever you get your podcasts. Be sure to subscribe to the Wellable Weekly newsletter. Thank you.
-
Share
-
-
-
-
