The rejection email arrives fast. Sometimes within hours of submitting an application, sometimes minutes after finishing a recorded video interview that felt more like a software demo than a real conversation. If you’re sitting in Minnesota wondering whether a person ever looked at your materials, the answer is often no. Automated resume screeners, video analysis tools, and personality assessment platforms now filter millions of applicants before a human recruiter gets involved.
The assumption that follows is understandable: if a computer decided, there’s nothing to be done. That assumption is wrong. Nichols Kaster PLLP has spent over 50 years representing employees and applicants against powerful institutions, and our team has earned First Tier recognition from U.S. News & World Report for labor and employment litigation. Minnesota’s civil rights framework already reaches algorithmic decisions, and a growing body of federal enforcement activity confirms that AI isn’t a legal shield for discrimination.
How AI Hiring Tools Produce Discrimination
Automated hiring tools don’t need to be programmed with bias to produce it. They learn from historical data, and when that data reflects decades of discriminatory hiring patterns, the algorithm encodes those patterns into its scoring. Amazon famously shut down an internal recruiting tool after discovering it was systematically downgrading resumes that included the word “women’s” and penalizing graduates of all-women’s colleges. The model had trained on a 10-year dataset of male-dominated hiring data and treated maleness as a proxy for job fitness.
The scale of the problem extends well beyond a single company. Research tracking 3.4 million applicants across 4 million job applications to 1,700 postings found that 26% of Black applicants and 15% of Asian applicants applied to positions where the AI system discriminated against their racial group. Those aren’t theoretical disparities. They represent real rejections.
The discrimination doesn’t only run along racial lines. Consider what some of these tools actually measure:
- Automated resume screeners and chatbots filter on keywords or knockout questions, eliminating applicants who don’t phrase their experience the way the algorithm expects.
- Video interview analysis platforms score eye contact, speech cadence, and facial expressions. These criteria can systematically disadvantage autistic applicants or people with speech-related disabilities.
- Game-style assessments and timed cognitive tests can penalize applicants with motor or cognitive impairments who are fully capable of performing the job.
In 2023, iTutorGroup paid $365,000 to settle an EEOC lawsuit arising from an AI candidate selection tool that had automatically rejected applicants based on age. The tool was doing exactly what it was configured to do. That’s the point.
Minnesota Law Doesn’t Care Whether a Human or an Algorithm Made the Call
The Minnesota Human Rights Act (MHRA) makes it an unfair employment practice to “refuse to hire” or “maintain a system of employment” that unreasonably excludes an applicant on a protected basis. That language comes from Minn. Stat. § 363A.08, subd. 2. A screening algorithm is a system of employment. Courts applying similar statutory language have treated automated tools as agents of the employer, and a federal judge expanded the Mobley v. Workday class action (which alleges Workday’s AI screening tools discriminate based on race, age, and disability) to include applicants rejected by the HiredScore AI platform on exactly that reasoning.
Two features of the MHRA matter especially for Minnesota applicants. First, it identifies thirteen protected classes, including public assistance status and local human rights commission activity, a broader list than federal Title VII. Second, it applies to employers of any size, even those with a single employee. Title VII only reaches employers with 15 or more, meaning a small employer could be immune to a federal discrimination claim but fully exposed under state law.
Disparate impact is the legal doctrine that covers facially neutral practices producing discriminatory outcomes. A plaintiff doesn’t need to prove the employer intended to discriminate, only that a practice had a statistically significant negative effect on a protected class. Once that showing is made, the burden shifts to the employer to demonstrate the practice is job-related and consistent with business necessity. An AI tool that screens out a disproportionate share of applicants with disabilities or applicants of a particular race faces that same analysis. Claiming the algorithm made the call doesn’t satisfy the business necessity defense.
A Minnesota Development Most Job Seekers Don’t Know About
The Minnesota Consumer Data Privacy Act (MCDPA), codified at Chapter 325M, took effect July 31, 2025. It gives Minnesota residents the right to question the results of automated profiling used to make decisions with legal or similarly significant effects, including employment opportunities. That right applies to individuals acting in a personal capacity. Note that the MCDPA defines “consumer” to exclude persons acting in a commercial or employment context, so the extent to which job applicants can directly invoke its protections isn’t yet fully settled.
Even so, the MCDPA matters beyond any direct right it may create for applicants. Violations of its automated-decision provisions can carry civil penalties of up to $7,500 per violation, enforceable by the Minnesota Attorney General, adding a separate accountability mechanism alongside the MHRA and Title VII for opaque algorithmic screening. A bill titled the Automated Decision Systems in Employment Act (HF 4445) is currently pending in the Minnesota Legislature and would add more direct worker protections. Minnesota hasn’t enacted it yet, but momentum is building at both the state and federal level.
What to Do If You Suspect an AI Tool Screened You Out
The evidentiary record you build now determines what options you’ll have later. Start documenting immediately.
- Save the job posting before it’s taken down, including the full text and any stated qualifications.
- Keep all rejection emails with their timestamps intact. An unusually fast rejection, minutes or hours after submitting materials, is itself evidence that no human reviewed your application.
- Note the name of any platform or automated tool referenced during the process, whether in an email, a login screen, or the application portal itself.
Deadlines are strict. A charge must be filed with the Minnesota Department of Human Rights within 365 days of the discriminatory act. A federal charge must be filed with the EEOC within 300 days. Minnesota is a deferral state, which is why the federal window extends to 300 days rather than the standard 180 days cited for other states. At least one widely shared employee-rights resource incorrectly tells Minnesota applicants they have only 180 days for the EEOC filing.
As of October 1, 2025, the MDHR and EEOC no longer automatically cross-file charges with each other. Preserving both your state MHRA claim and your federal claim now requires filing independently with each agency within its own deadline. Missing the EEOC filing doesn’t extend your MDHR window, and vice versa. Both filings are necessary to keep both options open.
Minnesota hasn’t passed an AI-specific hiring statute, but it doesn’t need one for your claim to move forward. The MHRA’s system-of-employment language covers algorithms, and federal enforcement through the EEOC has already produced settlements arising from AI hiring tools. If you applied for a position in Minnesota and believe an automated system may have screened you out on the basis of race, disability, age, sex, or another protected characteristic, our attorneys at Nichols Kaster PLLP can evaluate what you experienced. You can reach us at (877) 344-4628.