All employers must have a workplace harassment prevention policy. In many organizations, it’s been reviewed by a lawyer, approved by management, posted somewhere employees can find it and incorporated into orientation or annual training. HR can point to it when a complaint comes in. Managers know, at least in general terms, that harassment isn’t tolerated and that concerns should be reported. They also recognize that failure to carry out the policy can result in OHS penalties, civil lawsuits, and workers’ compensation claims.

But liability for not keeping the workplace harassment-free doesn’t end there.

The WestJet class action should make HR leaders look at their organization’s liability risks a little differently.

In August 2026, the Supreme Court of British Columbia approved a $4.5 million settlement in Lewis v. WestJet Airlines Ltd., a class action involving thousands of current and former female WestJet flight attendants who claimed they were harassed at work. The case had been moving through the courts for roughly a decade and was headed toward a lengthy trial when the parties reached a settlement.

What makes the case important isn’t just the money or the fact that it was brought as a class action. The significance for employers is that it opened a new door for employees to assert harassment claims against their companies: breach of contract.

The plaintiffs alleged that WestJet had made what became known in the litigation as an “Anti-Harassment Promise.” In very simple terms, the argument was that WestJet had made an implicit promise to flight attendants to ensure them a harassment-free workplace via implementation of an effective workplace harassment program that included, among other things, appropriate processes for reporting, investigating, and responding to harassment. That promise formed part of their employment relationship and was thus part of the employment contract, the plaintiffs argued.

The theory of the case was that WestJet had to pay the flight attendants who suffered harassment damages for violating that promise. They alleged that WestJet hadn’t simply mishandled individual complaints but systematically failed to provide the anti-harassment program it had promised its employees and had saved money by doing so.

That changed the nature of the case.

Instead of requiring thousands of employees to establish individual incidents of harassment and prove their personal damages one by one, the plaintiffs characterized the alleged failure as something that affected the entire class. Every employee who had been promised the program, they argued, had an interest in whether that program was actually provided.

The case settled before those arguments were tested at trial. But WestJet spent years trying to prevent the claim from proceeding, and the litigation survived long enough to become a certified class action involving thousands of employees.

The takeaway for companies and their HR directors: A workplace harassment prevention policy that may start life as an HR document required by OHS laws may also constitute a contractual promise. To the extent they represent an employment promise, the measures set out in the workplace harassment prevention policy also become contractual obligations. That’s a crucial distinction because workers aren’t allowed to sue their employers for OHS violations; their recourse is to seek psychological injury benefits under workers’ compensation. By contrast, employees can bring civil lawsuits for money damages, including class actions, against their employers for breaches of the employment contract.

Lesson 1. Read your harassment policy as if an employee’s lawyer wrote the questions

HR usually reviews a policy by asking whether it covers the right legal requirements. Does it prohibit harassment? Does it provide a reporting mechanism? Does it address retaliation? Does it describe the investigation process?

Those are necessary questions, but Lewis suggests employers should add another one.

What are we promising?

There’s a difference between saying an organization has a process for investigating harassment complaints and saying every complaint “will be fully investigated immediately.” There’s a difference between committing to reasonable confidentiality and promising complete confidentiality. There’s a difference between saying retaliation is prohibited and creating an impression that nothing unpleasant can ever happen to an employee after raising a concern.

Policy language tends to become stronger over time because stronger language sounds reassuring. Someone wants employees to know the company takes harassment seriously, so “we take appropriate action” becomes “we will take immediate action.” Someone wants employees to feel comfortable reporting a concern, so “information will be shared only where necessary” becomes “all complaints are completely confidential.”

Nobody is trying to create legal exposure. They’re trying to write a good policy.

The problem is that five years later, when an investigation is challenged, those words are no longer being read as reassuring HR language. They’re being compared with what the company actually did.

The WestJet litigation didn’t establish that every statement in a workplace policy is automatically an employment contract. It didn’t need to. The warning for employers is that plaintiffs were able to advance a contractual theory based on the commitments the employer allegedly made to employees.

That should be enough to change how HR reads its own policies.

Don’t strip useful commitments out of them. Employees deserve clear protections. But don’t casually promise something the organization isn’t capable of consistently delivering either.

Lesson 2. A good policy can’t rescue a bad system

Most organizations can produce a harassment policy within a few minutes if someone asks for it.

Finding evidence that the policy works is harder.

Suppose an employee reports harassment to a supervisor. Does the supervisor recognize that it needs to be escalated, or does the supervisor try to deal with it informally? If it reaches HR, how quickly is someone assigned to investigate? Is that person trained? Is the investigator sufficiently independent when the allegation involves a senior manager? Are witnesses interviewed properly? Are credibility issues documented? Does someone check on the complainant afterward? If retaliation starts two months later through scheduling, workload, performance management or exclusion, would anybody recognize the connection?

Those questions don’t show up on the policy itself.

They show up in the organization’s behaviour.

This is one reason the non-monetary part of the WestJet settlement deserves more attention than it has received. WestJet agreed to fund an independent assessment examining not only workplace harassment, but also the extent of underreporting and the effectiveness of the company’s reporting and response systems.

That is a very different exercise from asking whether a policy exists.

It asks whether people use the system, whether they trust it and whether it produces the results the organization says it will produce.

HR departments should be asking the same questions before a court, regulator or plaintiff’s lawyer asks them.

Lesson 3. One poorly handled complaint is a problem. A pattern is something else entirely.

Harassment investigations are often managed as individual cases.

Complaint comes in. File is opened. Interviews happen. Findings are made. Corrective action is considered. The file is closed and everyone moves on.

That makes operational sense, but it can create a blind spot.

What happens if the same manager appears in three complaints over four years? What if employees from one department repeatedly resign after raising concerns? What if complaints involving senior leaders consistently take twice as long to investigate? What if one location reports almost no harassment while exit interviews from the same location repeatedly mention bullying, inappropriate comments or fear of management?

You probably won’t see those things by reviewing one file.

You see them when you put the files beside each other.

The WestJet claim was dangerous precisely because the plaintiffs weren’t simply saying one complaint had been mishandled. They alleged a systemic failure. Their theory was that the anti-harassment program itself hadn’t been delivered as promised.

Whether that argument would have succeeded at trial will never be known. The case settled.

But the risk it illustrates is real.

HR needs to manage individual complaints properly, but someone also needs to periodically stand back and look at the system. Complaint numbers, findings, locations, departments, repeat respondents, investigation times, turnover following complaints and corrective actions should be reviewed for patterns.

Most organizations already possess much of this information. They just don’t put it together.

Lesson 4. Few harassment complaints may be good news. They may also be a warning.

This is one of the easiest mistakes for management to make.

The company receives two harassment complaints in a year and concludes that workplace harassment isn’t a significant problem.

Maybe that’s true.

Or maybe employees don’t believe reporting is worth the risk.

Harassment is chronically underreported for reasons that are easy to understand. Employees worry that their manager will find out. They don’t want to become known as difficult. They fear losing shifts, assignments, promotions or career opportunities. Sometimes they don’t think the behaviour is serious enough to justify a formal complaint. Sometimes they’ve watched someone else report a problem and decided they don’t want the same experience.

A low complaint rate tells you how many complaints were reported. It doesn’t necessarily tell you how much harassment occurred.

That distinction appears directly in the WestJet settlement. The independent assessment is required to consider the prevalence of harassment and the extent of underreporting. In other words, the settlement recognizes that counting formal complaints isn’t enough to tell you whether the reporting system is functioning.

HR should be doing the same thing.

Anonymous employee surveys can help. So can exit interviews, ethics hotline information, turnover data, employee-relations files and patterns in absenteeism or transfers. None of these proves harassment occurred, but together they can show where HR should look more closely.

There’s also a simpler question worth asking employees:

Do you believe you could report harassment here without hurting your career?

The answer may tell you more than the complaint log.

Lesson 5. Your investigation files may someday be read as a group

One investigation file can look perfectly reasonable on its own.

Put 50 of them together and a very different picture can emerge.

That’s worth remembering whenever someone writes an investigation note, delays an interview, decides not to speak with a witness or closes a complaint without explaining why.

The WestJet litigation became enormous. Thousands of documents were exchanged. The parties conducted extensive discovery. Numerous experts were retained, including 11 expert reports filed on behalf of the plaintiff. The court described the litigation as hard fought at every turn.

That’s what systemic litigation looks like.

Documents that were created years earlier to manage one workplace incident suddenly become part of a much larger evidentiary record. Lawyers compare them. Experts examine them. Patterns that weren’t visible when the individual documents were created become easier to see.

This doesn’t mean HR should write every investigation note defensively. That usually makes documentation worse.

It means investigation practices need to be consistent enough that HR would be comfortable having someone compare dozens of files.

Why did this investigation take four weeks while another took six months? Why was one complainant given regular updates while another heard nothing? Why was an external investigator retained when a junior employee complained about a director, but HR investigated internally when the allegation involved the vice-president?

There may be perfectly legitimate answers. The problem starts when nobody recorded them.

Lesson 6. Finding a problem is not the worst thing that can happen. Leaving it there is.

There’s an understandable instinct inside organizations to become cautious when an audit reveals a weakness.

If we change the procedure, will it look like we’re admitting the old one was bad?

That can lead to a strange form of paralysis where everyone knows something needs fixing, but nobody wants to create a document acknowledging the problem.

The WestJet decision offers a useful counterpoint.

After the litigation began, WestJet retained Ernst & Young to review aspects of its processes. According to Justice Hughes, issues were identified and WestJet made material improvements to its procedures. Those changes became relevant later because they made it more difficult for the plaintiffs to establish an ongoing breach during the latter part of the class period.

That is a practical lesson HR can use.

A problem discovered in 2026 may create questions about what happened in 2025. Ignoring the problem can create the same questions about 2027, 2028 and 2029.

Fix it.

Document what was changed and why. Train the people affected by the change. Set a date to check whether the new process is working. If it isn’t, change it again.

A functioning compliance program is supposed to improve when weaknesses are found. Pretending there was never a weakness is rarely a better strategy.

Lesson 7. Audit the promises scattered across HR, not just the harassment policy

The most interesting implication of Lewis may not be limited to harassment at all.

Employers make promises everywhere.

They appear in codes of conduct, accommodation policies, whistleblower procedures, investigation protocols, recruitment materials, employee handbooks, remote-work policies, performance management guides and training programs.

Sometimes different parts of the organization make different promises without realizing it.

The employment agreement may say one thing. The handbook says another. The careers website describes the culture differently. A manager adds another assurance during recruitment. The annual harassment training tells employees that “every report is investigated,” while the written policy gives HR discretion to determine whether a formal investigation is required.

Most of these inconsistencies are created innocently.

They can still become a problem.

That’s why policy review needs to move beyond proofreading and legal updates. HR should periodically compare what the organization tells employees with what the organization can genuinely and consistently do.

Start with the strongest words.

• Always.
• Never.
• Immediately.
• Guaranteed.
• Confidential.
• Every.

Those are the words most likely to turn an aspiration into a promise.

What WestJet actually settled

There’s a temptation with a $4.5 million headline to simplify the case too far, so it’s worth being precise about what happened.

The certified class initially included 3,593 current and former female WestJet flight attendants employed during the relevant class period. After 135 people opted out, 3,458 class members remained.

Under the settlement approved by the Supreme Court of British Columbia, WestJet agreed to pay $4.5 million without admitting liability. That amount also has to cover legal fees, litigation expenses, taxes, administration and other court-approved payments. The court estimated that roughly $1.61 million would remain for distribution to class members, which works out to approximately $465 per eligible employee if all eligible members participate.

Some class members objected to the settlement, including concerns about the amount ultimately available to individual employees. The court nevertheless concluded that the settlement was fair, reasonable and in the best interests of the class given the risks of continuing the litigation.

Those risks were significant.

There were competing expert assessments of the financial benefit WestJet had allegedly received. WestJet’s expert put the relevant cost savings in the range of approximately $2.5 million to $5.4 million. The plaintiff’s expert estimated approximately $27 million. Justice Hughes identified difficulties with the higher calculation and noted the uncertainty surrounding whether the plaintiffs would ultimately succeed with their disgorgement theory at all.

The $4.5 million settlement therefore shouldn’t be confused with a judicial finding that WestJet had saved $4.5 million by failing to prevent harassment.

There was no such finding.

The settlement was a negotiated resolution of litigation carrying substantial risk for both sides.

What this case does not say

Lewis needs to be used carefully.

The court did not decide that WestJet systematically breached its harassment obligations. It did not find that thousands of WestJet employees were sexually harassed. It did not hold that every Canadian workplace harassment policy forms part of an employee’s employment contract. And it did not order WestJet to pay $4.5 million after a trial.

WestJet expressly settled without admitting liability.

Those distinctions are not footnotes. They’re essential to understanding the case.

At the same time, employers shouldn’t use them as a reason to dismiss what happened.

The claim survived years of litigation. A contractual theory built around an alleged workplace policy promise was certified. Thousands of employees became members of the class. Both sides retained experts and prepared for a lengthy trial. The litigation eventually ended with a multimillion-dollar settlement and an independent assessment of WestJet’s harassment reporting and response systems.

You don’t need a final judgment against the employer for that sequence of events to teach something useful.

The HR reality check

There’s a fairly straightforward exercise HR teams can run after reading Lewis.

Take your harassment policy out.

Then take out your investigation procedure, employee handbook, code of conduct, harassment training materials and whatever managers receive when they’re trained to respond to complaints.

Put them beside each other.

Now look for promises.

What do you say happens when someone reports harassment? What do you say about confidentiality? How quickly do you say the organization will respond? What protection do you promise against retaliation? Who can employees report concerns to? What happens when the allegation involves the person they would ordinarily report to?

Then compare those statements with the last five or ten real cases.

Not the model investigation.

The real ones.

How long did they take? Who investigated them? Did employees receive updates? Was retaliation considered afterward? Were corrective actions documented? Did similar cases produce similar responses?

The gaps between those two versions of the organization, the one described in the policy and the one visible in the files, are where the risk sits.

The bigger lesson from WestJet

The $4.5 million figure will get most of the attention.

HR leaders should pay more attention to the legal theory that got the case that far.

The plaintiffs tried to take something employers have traditionally thought of as policy language and characterize it as part of the employment bargain. They argued that WestJet had promised an effective anti-harassment program and that the alleged failure to provide that program affected thousands of employees collectively.

The courts never decided whether that argument would ultimately succeed.

WestJet settled first.

But from an HR risk perspective, waiting for another employer to take the argument all the way through trial before paying attention would miss the point.

Workplace policies aren’t just documents employees are expected to follow. They also describe what employees can expect from the employer.

That makes the words important. More importantly, it makes the gap between those words and what happens in practice important.

The best protection isn’t softer policy language or fewer promises. It’s making sure the organization can stand behind the promises it chooses to make.

HRInsider Key Takeaways

WestJet’s $4.5 million settlement doesn’t establish that harassment policies automatically become employment contracts. It does show that workplace policy commitments can support much broader legal arguments than many employers expect.

The practical risk isn’t simply whether an organization has a compliant harassment policy. Employers also need to know whether reporting mechanisms work, employees trust them, investigations are timely and consistent, retaliation is identified, systemic patterns are reviewed and weaknesses are corrected when they’re found.

HR should also review the language used across policies, handbooks, training and employee communications. The question is no longer just whether the words comply with the law.

It’s whether the organization can deliver what those words promise.

Audit whether your organization can deliver its policy promises by downloaded the free Harassment Program Reality Check tool below.