What Happens when AI Hiring Platforms Completely Ignore the Human Experience?

The Rectech episode spotlights NOMA, a startup that inserts a disability‑focused trust layer into AI hiring platforms. Host Chris Russell and founder Tarn Ellis discuss how most conversational agents and assessment tools are built for normative users, leaving deaf, blind, or motor‑impaired candidates at risk. The conversation cites the recent HireVue litigation in Colorado, where captioning failures led to a deaf applicant being told to improve listening skills—a clear example of hidden bias causing legal exposure. By testing these systems with real disability users, organizations can surface gaps before they become costly lawsuits.

NOMA’s methodology begins with a discovery assessment that maps disability classifications onto every data point of a tool. Over six to eight weeks, a strike‑team of contractors from the disability community runs through the system, documenting escalations, remediation needs, and human‑impact metrics. Findings are compiled into a dashboard, video captures, and a comprehensive report linked to the NOMA Trust Index, which scores equity, accessibility, reliability, and transparency. While a formal badge is still under development, the company already issues rigorous certifications ranging from 177 to 300 sequential test steps, ensuring depth beyond algorithmic bias scans.

The market incentive is clear: roughly two to three billion people worldwide live with a disability, representing an untapped talent pool. Relying on synthetic data, as many frontier LLMs do, perpetuates exclusion and amplifies bias across multi‑tenant hiring platforms. Companies that integrate authentic lived‑experience testing not only reduce legal risk but also gain a competitive edge by demonstrating inclusive hiring practices. HR leaders should now demand evidence of disability testing, audit frequency, and clear accountability from vendors. Doing so transforms AI from a shiny shortcut into a responsible, equitable recruitment engine.