You turned on the AI tool because it was supposed to save time. Maybe it screens resumes. Maybe it handles scheduling. Maybe it tracks performance. You set it up, it ran, and you moved on. Here's the problem: that tool is making decisions about your people right now, and most business owners have no idea what it's actually doing.
That's not just a tech problem. It's a legal one.
AI Just Became the Biggest HR Compliance Risk in the Country
In May 2026, law firm Littler Mendelson released its 14th Annual Employer Survey, drawing on responses from more than 300 C-suite executives, in-house lawyers, and HR professionals. The headline finding: 84% of employers expect AI-related policy and regulatory changes to impact their business in the next year. That number doubled from 42% just one year ago.
AI knocked immigration and DEI off the top spot. That tells you how fast this is moving.
At the same time, 54% of those same employers said they're already using AI specifically for HR functions. Hiring, performance management, scheduling, you name it. The tools are in use. The guardrails are not.
Using the Tool Doesn't Mean You're Off the Hook
Here's where small business owners get caught off guard. You buy an HR software platform. The vendor says it uses AI to streamline hiring or manage performance. You take their word for it and move on.
What nobody tells you is that outsourcing the decision doesn't outsource the liability.
The EEOC has made clear that employers remain fully responsible under Title VII when AI-driven tools produce discriminatory outcomes. If an algorithm screens out candidates in a way that negatively impacts a protected class — by age, race, disability, gender — the liability attaches to you, not just the vendor. It doesn't matter that you didn't write the code.
A case already moving through federal courts, Mobley v. Workday, is testing exactly this theory. The court allowed the case to proceed on the argument that an HR technology vendor can be treated as an "agent" performing hiring functions on behalf of employers — which means both the platform and the businesses using it could face liability. That case is a signal. Courts are paying attention to what these tools do.
The Scale Problem Small Businesses Underestimate
Here's what makes AI discrimination risk different from a bad hiring decision made by a manager. A human makes a biased call, it affects one candidate. A biased algorithm makes that same call automatically, at scale, across every single applicant who comes through your system.
A bad algorithm doesn't cost you one complaint. It could cost you a class action.
This is especially true for fractional HR and HR outsourcing clients who rely on shared platforms without knowing how those platforms make decisions. If you're a business with 20 to 200 employees and your hiring or HR software uses any kind of automated scoring, filtering, or ranking, this applies to you.
Most Employers Still Don't Have a Written AI Policy
The Littler survey also found that while many organizations are moving to restrict what information employees can enter into AI tools (54%) or are building formal review processes (55%), a significant number are still playing catch-up on governance. Employers are adopting AI faster than they're building the policies to manage it.
That gap is where lawsuits live.
And at the state level, the rules are tightening fast. Colorado's AI Act takes effect June 30, 2026, requiring employers using high-risk AI tools in employment decisions to conduct bias testing, provide transparency notices to candidates and employees, and ensure that no final employment decision is fully automated. Other states are watching and moving in the same direction.
Three Things to Do Before You Get Caught Off Guard
You don't need a legal team or a compliance department to start getting this right. You need to take three steps this week:
Find out exactly what your HR software is doing. Call your vendor. Ask directly: does this platform use AI to screen, rank, score, or filter candidates or employees? Get it in writing. If they can't explain it clearly, that's a red flag.
Put a human on every final decision. No hiring decision, no termination, no disciplinary action should be executed purely by an algorithm. A real person needs to review and approve. Document that review.
Write down your AI use policy. Even a one-page document that describes what tools you use, what decisions they inform, and what human oversight exists puts you in a far better position than having nothing on paper at all.
AI in the workplace isn't going away. But neither is your obligation to manage it. The businesses that treat this as a compliance issue now are the ones that won't be explaining themselves to a lawyer later.
What AI tools are running in your HR process right now — and do you actually know what they're deciding?
Want help getting your AI use policies in place before they become a liability? Schedule a free 30-minute discovery call with Ricky Baez at baezco.com/contact-us.