AI Hiring Bias Litigation: Sirius XM Wins Dismissal — Plaintiffs Must Show How the Algorithm Discriminated
On September 30, 2026, Sirius XM was granted summary judgment in Harper v. Sirius XM Radio, LLC (E.D. Mich.), defeating a proposed class action alleging its iCIMS AI-powered hiring software "intentionally and disproportionately" discriminates against African American applicants (HR Dive). Applicant Arshon Harper applied to about 150 roles between November 2023 and November 2024 and used multiple email addresses to avoid "suspected algorithmic penalties for repeat applications." The court dismissed because the plaintiff "had not shown how the employer's alleged use of AI could have discriminated against him based on his race" and "failed to present both direct and circumstantial evidence" of discriminatory intent. Bloomberg Law's gloss: Harper "failed to establish a viable claim of employment discrimination, regardless of whether the company actually uses the artificial intelligence tools that he alleged."
iCIMS's statement after the ruling: "ICIMS develops AI under a responsible AI program with governance and safeguards designed to help mitigate potential bias. Our customers decide whether and how to use AI features, and their hiring teams make the hiring decisions."
The ruling lands in a shifting liability landscape. Federal disparate-impact enforcement has been pulled back — President Trump's April 2025 executive order directed agencies including the EEOC to cease enforcing disparate-impact liability, and in June 2026 the DOJ deemed the EEOC's disparate-impact guidelines unconstitutional — while fourteen state AGs simultaneously reaffirmed that they will enforce disparate-impact theories against algorithmic hiring, lending and tenant-screening tools (see 14 State AGs Reaffirm Disparate Impact Enforcement — With Algorithmic Hiring, Lending and Tenant-Screening Tools Squarely in Scope). Private litigation continues against vendors: the ADEA collective action against Workday (Mobley v. Workday) survived earlier dismissal attempts, and Workday "couldn't beat the state discrimination claims" under California's FEHA, with that case still ongoing. Harper shows the emerging defense playbook: courts require plaintiffs to specify the mechanism of algorithmic discrimination, not just allege AI was in the pipeline.