An AI Hiring Platform Secretly Built "Dossiers" on Job Applicants and Sold Them to Employers. Now It Is Being Sued.

In January 2026, job applicants filed a first-of-its-kind lawsuit against Eightfold AI Inc. The allegation: the AI hiring platform unlawfully collected personal data from social media profiles and internet browsing activities, created "dossiers" on candidates, ranked their "likelihood of success," and sold these reports to employers, all without the candidates' knowledge or consent.

The applicants were never given an opportunity to review the dossiers. They were never told what data was collected. They were never informed how the AI ranked them. They applied for jobs and were assessed by a system they did not know existed, using data they did not know had been collected, based on criteria they could not see or challenge.

What Eightfold Allegedly Did

According to the lawsuit, Eightfold's AI tools collected data far beyond what candidates voluntarily submitted in their job applications. The platform allegedly scraped social media profiles, internet browsing history, and other digital footprints to build comprehensive candidate profiles.

These profiles were then processed by Eightfold's AI to generate a "likelihood of success" score for each candidate. The scores were sold to employers as part of Eightfold's hiring analytics platform. Employers used these scores to make hiring decisions: screening in candidates with high scores and screening out candidates with low scores.

The fundamental problem is not that AI was used in hiring. It is that candidates were assessed using data they did not provide, by a system they did not consent to, based on criteria they cannot review or dispute. When a human recruiter makes a hiring decision, the candidate can ask why they were rejected. When an AI system makes that decision based on scraped social media data and opaque algorithms, the candidate has no recourse.

The Workday Class Action

Eightfold is not the only AI hiring platform facing legal challenge. The Mobley v. Workday class action alleges that Workday's AI hiring tools systematically discriminate based on age, race, and disability.

The plaintiff, a Black man over 40 with anxiety and depression, applied to more than 100 jobs through Workday's platform and was rejected from every one. The lawsuit alleges that Workday's AI screening system produces discriminatory outcomes that violate civil rights protections, and that Workday, as the operator of the AI system, is responsible for those outcomes even if the discrimination was unintentional.

This case is particularly significant because it tests a legal question that has enormous implications for the recruitment industry: when an AI system produces discriminatory outcomes, who is liable? The employer who used the system? The vendor who built it? Both?

The Scale of AI Screening

These lawsuits are not about niche technology used by a handful of companies. AI screening has become the default in corporate hiring.

According to the World Economic Forum's 2025 data, 88% of companies now use AI in some form in their candidate screening process. This means the vast majority of job applicants are being assessed by AI systems, whether they know it or not.

The outcomes of these AI assessments are troubling. Nineteen percent of organisations using AI in hiring report that their tools have overlooked or screened out qualified applicants. Nearly one in five companies using AI screening admit that their AI is rejecting people who should have been hired.

And candidates know something is wrong, even if they cannot identify the specific cause. Only 26% of applicants trust AI to evaluate them fairly. Three-quarters of job seekers believe AI hiring tools are biased against them, and the data suggests they are right.

The Stanford Evidence

In October 2025, Stanford University published research that directly tested AI resume-screening tools for bias. The findings confirmed what candidates suspected.

AI resume-screening tools gave older male candidates higher ratings than female candidates and young candidates with identical qualifications. Same resume. Same experience. Same education. Same skills. Different scores based on demographic characteristics that should be irrelevant to job performance.

This is not a theoretical risk. It is a measured outcome from controlled testing by one of the world's leading research universities. The AI systems that 88% of companies use for hiring decisions are producing scores that vary based on the candidate's age and gender: precisely the characteristics that employment law prohibits using as decision criteria.

The Legal Landscape Ahead

The Eightfold and Workday lawsuits are likely the beginning, not the end, of legal scrutiny on AI hiring tools.

The EU AI Act, which takes full effect in August 2026, classifies AI used in recruitment as "high-risk." This means any company using AI in hiring within the EU or for EU-based roles will face mandatory bias testing, transparency disclosures, and human oversight requirements. Non-compliance penalties can reach 35 million euros or 7% of global annual turnover.

In the United States, the EEOC has signalled increased scrutiny of AI hiring tools. New York City already requires bias audits for automated employment decision tools. Illinois requires consent before using AI to analyse video interviews. The regulatory direction is clear: AI hiring tools will face increasing legal requirements to prove they are not discriminatory.

For companies using AI screening today, the legal risk is not hypothetical. The Stanford research, the Eightfold lawsuit, and the Workday class action collectively establish that AI hiring tools can and do produce discriminatory outcomes. Any company using these tools without adequate bias testing and human oversight is accumulating legal liability with every hiring decision.

What This Means for Recruitment Agencies

The AI hiring bias story creates three distinct implications for recruitment agencies.

First, the liability question. If your agency uses AI screening tools as part of your candidate assessment process, you may be legally responsible for discriminatory outcomes those tools produce. The fact that you purchased the tool from a vendor does not necessarily shield you from liability if the tool's outputs result in discriminatory hiring recommendations. Before deploying any AI screening tool, agencies should conduct independent bias testing and maintain human oversight at every decision point.

Second, the competitive positioning. If you are a recruitment agency that does not use AI screening tools, the Eightfold and Workday lawsuits hand you a powerful differentiator. "We actually look at every candidate" is no longer just a quality argument. It is a legal risk argument. Clients using AI screening tools are accumulating liability. Clients using human recruiters are not.

Third, the candidate trust advantage. With only 26% of candidates trusting AI to evaluate them fairly, recruitment agencies that can guarantee human review of every application have a candidate attraction advantage. The best candidates, the ones who are in demand and can be selective, may increasingly refuse to apply through AI-screened platforms. A recruitment agency that offers human evaluation is positioned to attract candidates that AI platforms are driving away.

The Bigger Picture

The recruitment industry adopted AI screening because it promised efficiency: process more candidates, faster, at lower cost. The Eightfold lawsuit reveals what that efficiency actually looks like in practice: secret dossiers, undisclosed data collection, opaque scoring algorithms, and no candidate recourse.

Faster hiring that rejects qualified candidates is not faster. It is just broken. Cheaper screening that produces discriminatory outcomes is not cheaper when the lawsuits arrive. More efficient assessment that candidates do not trust is not more efficient when the best candidates refuse to participate.

The recruitment industry built its reputation on human judgment: the ability to see past a resume and assess whether a candidate will actually succeed in a specific role, at a specific company, with a specific team. AI screening tools promised to replicate that judgment at scale. The lawsuits suggest they replicated the biases instead.