What the Workday lawsuit means for Canadian HR leaders
Derek Mobley applied for more than 100 positions with companies that used Workday's platform as their exclusive hiring gateway. Every application was rejected. Some rejections arrived within hours. One came back in 55 minutes.
Mobley — who is Black, over 40 and has anxiety and depression — filed a charge of discrimination with the US Equal Employment Opportunity Commission (EEOC) in 2021. He later filed a federal lawsuit alleging that Workday's AI screening tools had discriminated against him based on race, age, and disability.
The lawsuit, Mobley v. Workday, Inc., has now survived multiple attempts at dismissal. In June 2026, a federal judge in San Francisco ruled that Workday must face claims that its software violated California anti-discrimination law and federal disability protections.
The case is the first of its kind to broadly target the algorithmic decision-making underpinning AI screening software — and legal experts say it's one that should be on HR's radar, even in Canada.
"Employers should be proactive," says Calvin To, employment and labour lawyer at Spring Law in Toronto, "because the Workday lawsuit in the States is evidence that AI systems can be a point of liability in an organization."
"Even if discrimination isn't intentional, employers can still be on the hook for it."
What the case actually argues
In the case, Mobley does not argue simply that Workday's clients discriminated against him. He argues that Workday itself is liable — as an employment agency, an indirect employer, and an agent of the employers whose hiring it managed.
The EEOC filed an amicus brief in support of Mobley, outlining why it believes each theory applies. On the employment agency question, the EEOC argued that Workday's tools perform precisely the functions traditionally associated with employment agencies: they evaluate candidates' qualifications, make judgments about suitability, and either reject applicants or advance them for further consideration.
The platform's branded assessments and personality tests, the EEOC noted, are the algorithmic equivalent of pre-referral testing that courts have long associated with employment agency conduct.
On the indirect employer theory, the EEOC pointed to the degree of control Workday exercised over applicants' access to jobs. Where Workday's platform was the exclusive entry point for hiring, the EEOC argued, the company functioned as a gatekeeper — controlling whether any applicant ever reached a human decision-maker.
On the agent theory, the statute itself is relevant: under U.S. legislation, the definition of "employer" includes "any agent of" an employer. The EEOC argued that when employers delegate significant hiring functions to a third-party platform — including the authority to reject candidates — that platform becomes an agent subject to the same anti-discrimination obligations as the employer itself.
Workday has denied the allegations, saying its tools evaluate only job qualifications and do not make hiring decisions.
'The employer is ultimately responsible'
For Canadian HR professionals, the vendor-as-agent theory is the part of Mobley that demands attention.
"Even if a company is using an AI product to assist in the hiring process, the company is ultimately responsible… because it’s the employer that ultimately makes the decision,” says To.
“A company can't use an AI product and then rely on it as a defense against any kind of discriminatory practice that might result."
It's important to not assume that AI systems will, by default, produce perfect and unbiased outcomes, he says.
"They're trained on real-world data. Real-world data comes from an imperfect society.
“And if an AI system doesn't account for discrimination, doesn't account for bias, for fairness… for equity, then that could potentially result in it producing discriminatory outcomes."
The pattern is not hypothetical. Amazon scrapped a recruiting algorithm several years ago after discovering it systematically favoured men — not because discrimination was programmed in, but because the system trained on data from a predominantly male workforce reproduced those patterns at scale.
Disclosure requirement not enough
While we haven't seen any lawsuits of this kind yet in Canada, that doesn't mean that it couldn't happen here, says To, citing protections for workers.
“We have human rights legislation, we have employment standards legislation, employees have rights under the common law as well,” he says.
“So, it's important that organizations, when they're thinking about implementing AI… that they do it in a manner that minimizes their liability.”
Ontario became the first Canadian jurisdiction to require AI disclosure in hiring when its Employment Standards Act amendment took effect Jan. 1, 2026, requiring employers with 25 or more employees to disclose whether they use AI in the applicant selection process.
But just having the disclosure or a disclaimer isn't enough to absolve an employer of obligations under the law, says To.
"You can't remove yourself from your obligations under the Human Rights Code, for example, or under the Employment Standards Act. These are requirements that every employer has to abide by, regardless of whether they're using AI."
The disclosure requirement tells a rejected applicant that AI was used. It says nothing about how the system was built, whether it was tested for bias, or what recourse exists when it gets something wrong.
"Even though it's best practice now to make that disclosure, it doesn't protect [the employer] from liability," says To. "They have to make sure they're implementing AI properly."
If a complaint arrives
Claims of algorithmic discrimination are costly and time-consuming. Applicants who believe they were screened out on a prohibited ground can file with the Human Rights Tribunal or pursue an action in court.
"Even if a claim doesn't have any merit, it still takes a lot of time to work on it and a lot of money for an organization," says To. "That's why I think it's best for organizations to be proactive."
If an internal review reveals a genuine bias problem, he recommends contacting the software provider directly, conducting an internal audit and, where warranted, looking for alternative tools.
"If companies are trying to improve the diversity of their workforce or make their hiring process more fair and equitable, there are tools out there that can help them do that."
And what if a vendor they're using faces a lawsuit?
"Employers shouldn't assume that just because an AI vendor is allegedly producing discriminatory results, that the employer will be off the hook," says To. "In fact, they should assume the opposite — that they will be on the hook — and act accordingly."
The human review question
AI systems are there to help us, they're not there to replace us, says To.
"You want to make sure [you’re] not just rubber stamping whatever the AI system produces.”
According to the EEOC's amicus brief, Workday's platform was in many cases the exclusive point of entry for job opportunities, with the system making automated decisions to reject candidates before they ever reached a human decision-maker.
"A prospective employee can only advance in the hiring process if they get past the Workday platform's screening algorithms.”
To describes the ideal as a system that is transparent throughout — one that shows how it is ranking applicants, what criteria it is applying, and how it is making judgments at each stage — with a human actively engaged in the process and making the ultimate hiring decision.
"Employers want to make sure that if they're using AI to assist in the hiring process, that they don't let it make decisions for them. They're ultimately responsible for the decisions they make."
What due diligence actually looks like
There are several questions that HR professionals should have when it comes to AI use, according to To.
"Employers should ask the software vendor if and how they've accounted for fairness and equity, whether it has any safeguards around discrimination, whether it has been designed with equity in mind, and whether the results are transparent so that the employer can conduct audits and make sure that it's doing what it's supposed to do."
In December 2025, Accessibility Standards Canada published CAN-ASC-6.2:2025 — Accessible and Equitable Artificial Intelligence Systems — a National Standard under the Accessible Canada Act requiring employers to validate AI hiring tools for equitable performance across disability groups and provide human alternatives to automated decisions.
Employment lawyer Lorin MacDonald, writing in Canadian Lawyer in March 2026, described the standard as a practical tool.
"This standard has immediate utility on both sides of the bar. Lawyers advising employers on AI procurement should treat CAN-ASC-6.2 as a due diligence benchmark. Lawyers representing applicants should treat non-compliance as evidence of a failure to accommodate."
The key here is for companies to figure out ways to implement it to achieve the outcomes that they want to achieve, says To, “and minimize the risk.”
