An artificial intelligence tracking system can contribute to workplace discrimination when it treats applicants differently because of a protected characteristic. While the technology itself is not unlawful, liability depends on how the employer selects and uses it.
Employers use automated tools to scan résumés, rank applicants, analyze interviews and measure assessments. Although these systems may appear neutral, they can repeat patterns in historical hiring data or penalize candidates whose speech, disabilities, employment gap or backgrounds differ from those represented in the training data.
When does AI screening become discriminatory?
Federal employment laws apply even when software assists with a hiring decision. Discrimination may involve intentional unequal treatment. It may also result from a neutral screening practice that disproportionately excludes a protected group and is not related to the position or necessary for the business.
A résumé tool trained on records from a workforce dominated by one demographic could learn to favor similar applicants. Video software may inaccurately evaluate someone with a speech impairment. A system that rejects candidates based on graduation dates could create age-related concerns.
An employer cannot avoid responsibility because a vendor designed the software. Before deployment, employers should:
- Examine what information the tool collects
- How scores are generated
- Whether validation supports its use
- If results show unequal outcomes
They should also provide a reasonable accommodation when a disability prevents an applicant from completing an automated assessment in the standard format.
Applicants who suspect discrimination should preserve job postings, emails, assessment instructions, screenshots, rejection notices and accommodation requests. Evidence showing that qualified candidates from a protected group were consistently screened out may support further investigation with the help of a qualified legal practitioner.
AI does not create a separate discrimination claim. Instead, its use may become evidence that an employer violated federal, state or local employment protections.