Workplace AI Bias Just Cost Companies Millions in Lawsuits

If an algorithm decides you're not a culture fit, who do you sue? If your company uses AI to rate your performance, do you get to see what it's scoring? And if you use AI at work, are you improving your workflow, or handing your employer a reason to terminate you? I'm Mila Arutunian, an employment attorney, and this is Punching the Clock. I'm calling out bad behavior in the workplace and breaking down the legal implications. Let's make sure you have the legal tools to stand up for yourself. What if I told you that a computer might already be making decisions about your job, and you never know it? Not in a futuristic way, not in some tech company lab — right now, in everyday workplaces. Deciding whether your resume ever gets seen. Helping shape your performance review. Tracking how you work, minute by minute. And when something feels off, you're told, "It's not me, it's the system." Today we're talking about workplace AI, and what it's actually doing behind the scenes. Let's start with hiring, because this is where most people are already being impacted, whether they realize it or not. When you apply for a job online, your resume almost never goes straight to a human. It goes to software first. That software scans your resume, compares it to the job description, and ranks you against other applicants. But here's what people don't realize: those systems are trained on past data, and past data reflects past decisions. So if a company has historically hired people from certain schools, backgrounds, age groups or career paths, the software may quietly favor people who look like that and filter out everyone else. That can mean getting rejected because you changed industries, because you took time off to care for a child or a parent, because your resume doesn't use the right buzzwords — even though you're fully qualified. And the most frustrating part is that you'll probably never be told any of this. You'll just get automated emails: "We've decided to move forward with other candidates." No explanation, no feedback, no human judgment. Now let's talk about performance, because this is where things get really personal. More and more employers are using AI tools to help evaluate employees — not always to make final decisions, but to shape them. These tools might summarize your emails, track output, analyze response times, or generate draft performance reviews. So if you've ever read a performance review and thought, "That doesn't sound like my manager, or me" — there's a reason. AI doesn't understand context. It doesn't know you were mentoring a junior employee. It doesn't know that you handled a crisis that pulled you off your regular tasks. It doesn't know when your workload doubled, or when someone else quit. It just sees numbers, patterns, comparisons. And once something negative makes it into a performance file, even if it's wrong, it can follow you. It can affect promotions, your raises, who gets put on a performance improvement plan, who gets laid off first. And now let's talk about tracking, because this is the one that makes most people uncomfortable, especially with remote or hybrid work. Some employers use software that tracks your activity. That can include mouse movement, keyboard activity, screenshots, how long your computer is active. And here's the problem: activity is not always the same as productivity. You could be thinking, reading, on a call, working offline — but the system might flag you as inactive. And once you're flagged, you're on someone's radar, even if you're doing your job really well. Employees often don't know that tracking is happening, don't know what's being measured, and don't know how the data is being used. And when they're questioned about it, they're told, "The system shows us you weren't working." And that brings us to the core of the problem: when AI gets involved, accountability gets fuzzy. If AI screens people out, bias can hide inside the data. If AI evaluates performance, mistakes can snowball. If AI tracks activity, normal human behavior can look like misconduct. And when employees push back, they're often told that it's not a person doing this, it's just how the system works. Which is why courts are starting to pay attention. Here's a real example. If AI blocks you from 100-plus jobs, who do you sue? Let's get into some case files. This one is currently ongoing and paving the way to get more laws in the books regarding AI protections. Case file number one is Mobley v. Workday. If the employer says the software did it, and the software company says the employer did it, who's responsible? There's an ongoing lawsuit where a job applicant alleges Workday's AI-driven hiring tools screened him out repeatedly, raising discrimination claims. This individual applied to over a hundred jobs and kept getting rejected again and again. The common thread: the employers were using the same hiring software. He alleges discrimination based on race, age and disability. What matters isn't just his claim — it's what the court said next. A federal judge allowed the case to move forward, including claims that the software company itself could potentially share responsibility for the discrimination. So what's the translation? You don't get to hide behind an algorithm anymore. This case is really going to test whether an HR tech vendor can face liability like an employment agency does. I'll follow up on this one as it goes through court. But this isn't the only example. As AI rises, so do these lawsuits. In another case, the EEOC went after iTutorGroup. The federal government sued the company because its recruiting software automatically rejected older applicants. That case settled for real money, and the message was clear: civil rights laws still apply, even when decisions are automated. So the legal system is basically saying, you don't get to automate your way out of civil rights. But employers, unfortunately, still repeat the same mistakes. They say AI can't discriminate because it's not human. They say it's the vendor's fault. They say there aren't any rules yet, so we can kind of do what we want. But none of that is true. The law really looks at outcomes. Employers choose the tools, and rules are already rolling out at the city and state level. So now let's flip the camera, because this affects employees in another way. Number one: "AI can't discriminate because it's not human." This is a myth. If an AI tool creates disparate impact or reinforces bias, employers can still face exposure under traditional anti-discrimination frameworks such as Title VII, the ADA or others, depending on the facts. The EEOC has specifically focused on algorithmic decision tools and adverse impact analysis. Number two: "If a vendor built it, we're not responsible." This is a myth. The Workday litigation is actively testing vendor responsibility, but the safer assumption for employers is: you're still accountable for outcomes when you deploy tools in hiring and employment. Number three: "AI laws don't exist yet." That is a myth. They are here, and they are expanding. New York City has a new local law against AI bias, implementing notice requirements before using automated employment decision tools. California approved regulations aimed at preventing employment discrimination tied to automated systems and AI. Illinois amended its Human Rights Act, restricting discriminatory impact from AI in employment decisions, and notice expectations. Colorado also implemented the AI Act, leading to consumer-protection-style obligations for high-risk AI systems, including in employment contexts. Here's the thing: legislation is moving forward, so employers and employees need to keep up. The policy moment: can employees use AI at work? If you use AI to work faster, do you get rewarded, or do you get replaced? People ask me all the time, can I use AI at work? And here's the uncomfortable truth: something can still be legal and still get you fired. I've already seen employees disciplined and terminated for using AI the wrong way. Common mistakes include pasting confidential information into public AI tools, letting AI do work that requires your judgment, or relying on AI for decisions about co-workers. AI sounds confident, but that doesn't mean it's right. And it definitely doesn't mean your employer will protect you. So here's my rule of thumb. Using AI to organize your own notes, draft a rough email, or brainstorm ideas without sensitive information is usually okay. But uploading company documents, letting AI make final decisions, or using it for legal or HR conclusions — that's where people honestly get burned. So if your employer is using AI, or planning to, you're allowed to ask questions. What tools are approved? What data can't be shared? When does a human have to review decisions? Those aren't difficult employee questions — they're basic protections. Okay, guys, here's the punchline. AI does not make workplaces fair. It doesn't remove bias — it can scale it. This makes rules, transparency and accountability more important than ever. And when powerful tools are used without guardrails, employees are usually the first ones to pay the price. So I'll just leave you with this: if a machine makes a call, someone still owns the consequences. Please reach out to me if you think you've been treated unfairly because of AI workplace tools. And ask yourself: would you trust AI to decide who gets hired, or who gets fired? Has AI shown up in your job in a way that has made you uneasy? I'm Mila Arutunian, and this is Punching the Clock. Drop your workplace story in the comments, and I might cover it next.