Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Yes. An AI hiring tool can create discrimination risk when it makes or materially influences who is screened, assessed, or selected. Existing employment-discrimination rules still apply to those decisions. But the available official sources do not establish that AI hiring bias typically costs employers millions, or provide a standard loss amount; that figure is a risk framing, not a documented general outcome.
How AI hiring tools can create discrimination risk
Hiring tools may rank résumés, screen applications, assess tests or video interviews, or otherwise inform a selection decision. The legal concern is not limited to a system that makes the final choice: the Equal Employment Opportunity Commission (EEOC) says Title VII applies when automated systems make or inform employment-selection decisions. If a procedure disproportionately excludes people in a protected group, an employer may face a disparate-impact question even if the tool was not designed to discriminate. The EEOC’s FY2023 Agency Financial Report summarizes the agency’s technical assistance on these issues.
A score, ranking, or vendor claim that a tool is “objective” does not by itself establish that the tool measures job-related qualifications fairly. Employers need to consider the job-related basis for the assessment, its validity for the use at hand, selection outcomes, and whether people can complete it with reasonable accommodation.
Does the four-fifths rule prove a hiring tool is fair?
No. The four-fifths rule is a screening indicator used in the Uniform Guidelines on Employee Selection Procedures, not a safe harbor or conclusive legal test. The EEOC’s FY2023 report summarizes its warning that following the rule “does not guarantee that a procedure won’t be found to have a disparate impact on a prohibited basis.” A result that meets the guideline does not end the inquiry; employers still need to consider the selection procedure and the relevant evidence.
#1 Best Overall
What employment-AI rules apply in different places?
| Location | What the cited authority establishes | Important boundary |
|---|---|---|
| United States: federal baseline | Title VII applies when automated systems make or inform selection decisions, and disparate-impact analysis can apply. EEOC FY2023 Agency Financial Report | This is not an AI-only exception to existing employment law; the relevant facts and legal analysis depend on the employer’s use and the selection procedure. |
| New York City | Local Law 144 imposes requirements for covered automated employment decision tools (AEDTs): a bias audit within one year before use, public availability of audit information, and required notices. The Department of Consumer and Worker Protection (DCWP) says enforcement began July 5, 2023. NYC DCWP’s AEDT page | The law’s definitions and rules determine whether a tool is covered; it does not apply to every hiring algorithm. NYC311’s AEDT summary describes candidate and employee notices, accommodation instructions, and public information about data use and retention. |
| European Union | Regulation (EU) 2024/1689 classifies specified AI systems used for recruitment and selection, and certain decisions affecting work relationships, as high-risk. The regulation cites potential effects on employment prospects, livelihoods, workers’ rights, and the risk of perpetuating historical discrimination. The enacted EU AI Act | Classification and obligations depend on the system and its use. The regulation has implementation provisions; a particular employer should check the operative requirements and applicable timing rather than assume one deadline applies to all systems. |
This is not a complete inventory of state and local U.S. laws. Employers operating across jurisdictions should identify the rules currently applicable to each location and use case.
Why disability access matters in an automated assessment
EEOC and Department of Justice guidance highlights several distinct ADA risks. An assessment may screen out an applicant who could perform the job with or without reasonable accommodation; an employer may fail to provide an accommodation that enables a person to take the assessment; or a tool may elicit disability-related or medical information in ways that raise restrictions on inquiries and medical exams. These are risks to assess, not a finding that every automated test violates the ADA. See the agencies’ joint warning on disability discrimination and algorithmic tools.
Rank #2
An accessible process should make it possible for applicants to request an accommodation and for the employer to respond before the assessment unfairly screens them out. In New York City, required notices for covered tools include accommodation-related information; applicants elsewhere should not assume they have the same notice in the same format.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsWhat employers should check before and during use
Responsibility depends on jurisdiction, system design, and how the tool is used. A vendor audit or contract does not, by itself, establish that an employer has met its obligations. A practical review should follow the tool through the hiring process rather than treating an audit as a one-time approval.
- Identify the applicable rules. Map the locations where the tool is used and determine whether the tool and decision fall within specific requirements, including NYC’s AEDT law where relevant.
- Document the tool’s role. Record which hiring stage it affects, what inputs it uses, whether it ranks or filters applicants, and how its output influences a human decision.
- Establish job relevance and validity. Identify what job-related requirement the assessment measures and what evidence supports using it for that purpose.
- Review selection outcomes. Examine outcomes for relevant groups and investigate potential adverse impact. Do not treat the four-fifths rule as a guarantee of compliance.
- Provide an accommodation route. Tell applicants how to request an accommodation and ensure the assessment can be adapted where required.
- Meet notice, audit, and record duties. For covered NYC AEDTs, confirm the audit timing, public availability of audit information, and required notices. Preserve appropriate records and review data use and retention.
- Monitor changes and results. Reassess when the tool, job, applicant population, or hiring process changes, and check whether outcomes continue to support the tool’s intended use.
When comparing tools or assessment approaches, ask about the decision stage affected, job-related evidence, outcome monitoring, accommodation access, available audit and notice information, data retention, and support for independent review. These are useful investigation questions, not a universal scoring framework prescribed by the cited agencies.
What applicants can do if a tool affects their application
- Ask the employer whether an automated tool is used in the hiring process and which assessment or decision it affects.
- If a disability or access need could affect your ability to complete an assessment, ask how to request an accommodation before taking it.
- If you are applying in New York City and the tool is covered by Local Law 144, review the employer’s notice and the publicly available audit information; the law includes notice requirements for covered tools.
- Keep copies of relevant notices and communications, including your accommodation request and the employer’s response.
These steps can help clarify how the process works, but they do not guarantee a particular hiring outcome or establish that a tool violated the law.
Rank #4
What the “millions” claim does—and does not—mean
Discrimination exposure can be serious, but the cited official sources do not establish a typical multimillion-dollar loss, a general expected-loss figure, or how often AI hiring tools discriminate. A concrete amount would require evidence tied to a specific case, study, or legal outcome. The sound conclusion is narrower: automated hiring decisions can trigger existing legal duties, and weak review of job relevance, adverse impact, accommodation, or required notices can create compliance risk.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
