AI Decisions Are Triggering Lawsuits. Is Your SMB Covered?

Why are AI-related employee lawsuits suddenly a small business problem?

Small businesses using AI to screen resumes, schedule shifts, flag underperformers, or handle customer complaints are making legally consequential decisions, often without realizing it. When an employee is passed over, demoted, or let go and an algorithm was involved, the legal question is no longer hypothetical. Courts and regulators are actively working through what liability looks like, and plaintiffs' attorneys are paying attention.

This isn't a Fortune 500 problem anymore. The same tools that cost $50 a month on a SaaS plan can generate a six-figure lawsuit if they're used to make employment decisions without proper documentation, human oversight, or bias testing.

What kinds of AI use are actually creating legal exposure?

Three categories show up most often in early cases and regulatory guidance:

1. AI-assisted hiring and screening Using a tool that scores or ranks applicants based on resume language, video interview analysis, or skills assessments. Illinois passed the Artificial Intelligence Video Interview Act in 2020 requiring disclosure and consent. New York City's Local Law 144 mandates bias audits for automated employment decision tools. If you're using an ATS with AI scoring and you're in a covered jurisdiction, you may already be out of compliance.

2. AI-influenced layoffs or performance management If a tool surfaces who to cut and you act on it without documented human review, the resulting termination can look discriminatory even if no one intended it. Workforce reduction decisions tied to algorithmic output are exactly the scenario Counterpart's new Agentic Insurance platform was built to cover.

3. AI in customer service and communications Chatbots that deny claims, refuse refunds, or give incorrect information are generating consumer protection and ADA complaints. If your bot handles anything consequential, you own the output.

"We are witnessing a generation of small businesses deploy AI tools without understanding that they just became the decision-maker of record." (Counterpart, April 2026)

How big is the actual legal risk?

Counterpart's $50M Series C, bringing total funding to $106 million, was raised specifically because institutional investors see this as a large and underserved market. Valor Equity Partners led the round. That's not a charity investment. It reflects a thesis that AI-related employment litigation is growing fast enough to support a specialty insurer at scale.

Employment practices liability (EPL) claims are already expensive before AI is involved. The average EPL claim costs around $160,000 to defend and settle, according to Hiscox. Adding an algorithmic decision layer gives plaintiffs' attorneys another angle to pursue and another reason to go after discovery. Small businesses typically can't absorb that, and most standard business owner policies don't cover EPL at all.

Does existing insurance cover AI-related employment claims?

Almost certainly not in full. Standard EPL policies were written before generative AI and automated decision tools were widely deployed. They cover wrongful termination, discrimination, and harassment in the traditional sense. They are not written to address:

• Liability from a biased hiring algorithm
• Claims arising from AI-generated performance reviews
• Regulatory fines from non-compliant automated decision tools
• Content liability from AI-generated customer communications

This is the gap Counterpart is explicitly targeting with its Agentic Insurance product. Whether Counterpart is the right fit for your business is a conversation to have with your broker, but the more important immediate step is auditing what your current policies actually say about algorithmic decision-making.

What does good AI governance look like for a small business trying to reduce this exposure?

Governance doesn't have to be a 40-page policy document. For most SMBs, it comes down to four practical controls:

| Control | What it means in practice | Why it reduces risk | |---|---|---| | Human sign-off on AI outputs | No hire, fire, or denial decision executes without a named person reviewing and approving | Creates a documented decision chain that isn't purely algorithmic | | Bias logging | Keep records of what tool was used, what it recommended, and what the human decided | Gives you evidence that a human was in the loop | | Jurisdiction check | Know which states or cities your employees are in and whether AI hiring laws apply | Avoids the compliance violations that trigger regulatory action | | Policy disclosure | Tell employees if AI is used in performance management | Required in some jurisdictions, good practice everywhere |

None of this is exotic. It's basic documentation hygiene applied to a new category of business decision.

Should small businesses stop using AI in HR because of this risk?

No. That would be like stopping the use of spreadsheets because someone once made a bad financial decision with one. The tools are not the problem. Deploying them without process is the problem.

The businesses that will get hurt are the ones that plugged in an AI recruiting tool or a performance management platform, let it run autonomously, and have no documentation trail showing that humans reviewed the output. That's not a technology failure. That's a process failure, and it's fixable.

For most small businesses, the right answer is to keep using these tools, build a lightweight governance layer around them, check whether your EPL coverage addresses AI, and make sure you understand the employment laws in your operating jurisdictions.

What we'd actually do

• Audit your AI touchpoints in HR and customer service this week. List every tool that influences a hire, fire, promotion, demotion, or customer decision. If you can't list them, you can't govern them.
• Pull your EPL policy and read the exclusions. Specifically look for language around automated decision tools, algorithmic systems, or technology-assisted decisions. If it's silent on AI, call your broker before your next hiring cycle.
• Build a one-page human review log. For any consequential AI-assisted decision, require the responsible manager to document what the AI recommended and what they decided. This takes five minutes and dramatically changes your legal position if you're ever in discovery.