HR Tech · AI recruiting communications

HR tech company eliminated EEOC exposure in AI-generated recruiting communications

Published May 5, 2026

The problem

The customer provided AI-assisted recruiting tools to mid-market employers. Their compliance team identified a systemic risk: AI-generated job outreach and interview scheduling emails occasionally contained language that could constitute disparate impact discrimination under Title VII, the ADEA, or the ADA — references to age-related experience thresholds, physical requirement language, and cultural fit phrasing that courts have found discriminatory.

The risk

EEOC discrimination charges in AI recruiting are a growing area. The EEOC's Technical Assistance Document on AI (2023, updated 2025) explicitly states that employers are liable for discriminatory AI outputs regardless of whether a human reviewed them. Per-charge settlements average $40k–$150k; pattern-and-practice suits range into eight figures.

Implementation

Applied the DataVibe HR recruiting compliance bundle (built from EEOC guidance, OFCCP regulations, and state-level bias laws). The policy flags: age-proxying language (recent graduate, young and energetic, recent college graduates), physical requirement language without BFOQ documentation, cultural fit phrasing flagged in EEOC guidance, and language that implies preference based on protected class characteristics. BLOCK severity for direct protected-class references; WARN severity for proxies that require recruiter review.

What changed

• 43 discriminatory language patterns blocked across 2,847 recruiting emails.
• Zero EEOC complaints filed against any employer using the platform in the 8-month window.
• HR compliance report generated for two enterprise customer audits — full chain of evidence.
• Recruiter review queue averaged 34 seconds per item; 91% resolved without edit.