AI Hiring Assessments Are Almost Never Tested for Disability Bias — and One Vendor Just Admitted It on the Record
Not legal advice. This article discusses legal and regulatory developments and is generated with AI assistance under human editorial review. It is provided for informational purposes only and does not constitute legal advice. Laws and regulations vary by jurisdiction and are subject to change. You should consult your own qualified legal counsel before taking or refraining from any action based on the content of this article. Read our AI Transparency statement.
The News
Job Board Doctor published an investigation on 6 August revealing that AI-powered hiring assessments are almost never tested for disability bias — and got one vendor, Maki People, to confirm it on the record. When asked whether its Mochi assessment conducts adverse impact analysis by disability and veteran status, Maki People answered in full: it does not. The response is notable not because Maki is an outlier, but because it appears to be representative. Most AI hiring assessment vendors test for race and gender bias because those are the categories regulators historically scrutinise. Disability status — harder to measure, harder to self-report, easier to overlook — falls into a compliance gap that no one is enforcing with any vigour. Yet AI assessments that measure cognitive speed, response patterns, or verbal fluency may systematically disadvantage candidates with certain disabilities, and without testing, no one knows by how much.
Why It Matters
This matters now because AI pre-hire assessments are scaling faster than the compliance infrastructure around them. Vendors in this space have multiplied from roughly one in 2021 to more than three dozen today, and by mid-2026, an estimated 80% of high-volume recruiting starts with some form of AI-powered screen. If these tools aren't tested for disability bias, every hire made through them carries an unquantified legal risk — and an ethical one. The EU AI Act's high-risk obligations (now postponed to December 2027) would eventually require bias testing across protected characteristics, but in the US, the regulatory patchwork leaves disability bias monitoring largely to employer discretion. The EEOC's 2023 guidance on AI and the ADA made clear that tools which screen out individuals with disabilities violate federal law — but guidance without enforcement is a suggestion. TA leaders using these tools are operating in a compliance vacuum, and that vacuum is where class-action lawyers do their best work.
Taking Action
All action is situated in your corporate context. These are starting points, not prescriptions — choose what fits.
Connect
Who to talk with in your enterprise
Your HRIS or HR technology lead should request — in writing — adverse impact documentation from every AI assessment vendor you use, specifically asking for analysis by disability status. If the vendor can't produce it, that's your answer. Your legal or compliance team should review your AI hiring tool inventory against the EEOC's 2023 guidance on algorithmic fairness and the ADA. If you operate in the EU, your DPO should map every AI assessment tool against the EU AI Act's high-risk classification framework, even if enforcement has been delayed.
Inspect
What to take a closer look at
Audit every AI-powered assessment, screen, or filter in your hiring stack. For each one, document: (a) what it measures, (b) which protected characteristics it has been tested against, (c) which it hasn't, and (d) when the last bias audit was conducted. If the vendor can't answer (c) and (d), flag the tool as carrying unquantified legal risk. Then check your own data: are candidates who self-identify as having a disability progressing through AI-screened stages at the same rate as those who don't?
Reflect
What to think about more deeply
If you wouldn't deploy a recruitment process you couldn't defend in a tribunal, why are you deploying AI assessments you can't prove are fair?