The Rise of “Automated Ghosting” Claims
In 2026, “ghosting”—the practice of ignoring job applicants—is no longer just a breach of etiquette; it’s a legal trigger. When an AI agent rejects a candidate based on “proxy data” (like gaps in a resume that correlate with disability or caregiving), it can create a Disparate Impact claim. Because the candidate never speaks to a human, they often perceive the rejection as discriminatory, leading to costly class-action litigation.
The “Mobley v. Workday” Effect
A landmark 2026 legal trend is the expansion of liability from the employer to the AI vendor. However, US courts are increasingly holding the employer responsible for the “Agency” of their software. If your AI tool “ghosts” a protected group because it was trained on biased historical data, your Employment Practices Liability Insurance (EPLI) will be your only line of defense against “Nuclear Verdicts” that now frequently exceed $10 million.
New Mandatory Compliance for 2026
To secure an EPLI renewal in May 2026, many US insurers now require proof of:
- Bias Audits: Annual third-party testing of your hiring algorithms to ensure they aren’t screening out candidates based on age, race, or gender.
- The “Right to Explanation”: Following the lead of the Colorado AI Act, more states now require businesses to tell candidates why they were rejected if an automated system made the decision.
- Human-in-the-Loop (HITL): Insurers are offering lower premiums to firms that mandate a human “final look” before any candidate is officially rejected.
Protecting Your Business (The EPLI Strategy)
- Check Your Sub-Limits: Many standard EPLI policies have small caps for “Failure to Hire” claims. In 2026, you should look for a dedicated AI Liability Rider.
- Vendor Indemnification: Ensure your contract with your HR tech provider includes a “Fairness Guarantee” where they share the defense costs if their algorithm is found to be biased.