Your EPL Policy Wasn't Priced for a Screening Algorithm
A CFO opens the renewal binder this month. General liability, tabbed and reviewed. Cyber, tabbed and reviewed. Employment Practices Liability sits near the back, untouched since the broker walked it through three years ago, because nothing about a hiring policy ever seemed like it needed a second look.
It's not a hiring policy anymore. It's a software policy that happens to touch hiring, and carriers are now pricing that distinction directly. According to Insurance Business magazine (July 13, 2026), AI exclusions now appear in roughly 10% of the EPL market, with carriers diverging sharply on whether to write them at all. Travelers' Chris Williams put the underlying problem plainly in that reporting: "The vast majority of EPL policies do not make a distinction between wrongful employment practices committed by humans versus AI." Broker Lucas Roberts of Burns & Wilcox described some carriers "right out of the gate" carving the exposure out entirely.
The case forcing the issue is Mobley v. Workday, a proposed class action alleging Workday's AI screening tools discriminated by age, race, and disability. In June 2026, a federal judge rejected most of Workday's motion to dismiss, letting key California and federal disability claims proceed. That ruling is what turned an abstract underwriting question into a live one: if your AI screening tool is found to have discriminated, does your EPL policy respond, or does it treat that as a different, excluded risk entirely.
It's not a coverage question anymore. It's a documentation question. The policy language matters, but so does the paper trail behind it: can you show, in writing, whether your carrier's EPL policy distinguishes a human recruiter's judgment from a decision an algorithm made or materially assisted. Most policyholders have never asked. Most brokers, before this year, never needed a good answer.
If the answer comes back as a hard exclusion with no workaround, the market is already building the alternative. HSB, a Munich Re subsidiary, launched an affirmative AI Liability product in July for small and midsize businesses, covering third-party bodily injury, property damage, and personal or advertising injury claims tied to AI use (IA Magazine). Newer entrants like Corgi are underwriting specifically for AI errors and hallucination-driven losses (Marketplace). "Our carrier won't cover it" is no longer where this conversation has to end.
This is the same discipline this newsletter has been pushing for months, just extended one line further. The 30-Day AI Coverage and Contract Audit already asks three questions of every broker: has your carrier filed an AI exclusion endorsement in your states, does your policy respond to a claim from an AI failure, have you reviewed vendor contracts for indemnification language. As of this week, add a fourth: does our EPL policy distinguish between a human recruiter's decision and an algorithm's, and if so, exactly where is that line.
This month's action: get your broker's answer to that fourth question in writing before your next EPL renewal. If the answer reveals a gap, ask what the market now offers to close it. That conversation is far cheaper before a claim than after one.