#backgroundcheckservicescanada

Introduction

Social media has quietly become part of the hiring process at most Canadian companies. Surveys consistently show that around 70% of hiring managers have looked up a candidate online before making an offer. The problem isn’t whether it happens. It’s how.

When a hiring manager reviews a candidate’s profiles from their personal account, they expose the company to bias claims, privacy complaints, and human rights challenges. None of that goes away just because the information was technically public.

This guide walks through what Canadian employers can legally review, what they cannot, and how to run social media checks in a way that holds up to scrutiny.

What Is a Social Media Background Check?

A social media background check reviews a candidate’s public online presence to identify content that could pose a risk to the workplace. Done properly, it surfaces job-relevant signals that a criminal record check will not catch: hate speech, harassment, threats of violence, or evidence of drug use or other behaviour that could affect the role.

Done improperly, it captures everything else too: the candidate’s religion, family status, political views, health information, and a hundred other details that have no place in a hiring decision and create direct legal exposure if they do.

The difference between a useful check and a liability is structure.

Is Social Media Screening Legal in Canada?

Yes, with conditions. Canadian privacy law and human rights legislation set the boundaries. The general principles are straightforward:

The information has to be collected for a reasonable, job-related purpose. Candidates should be informed that the check is happening. The process has to be applied consistently and cannot be used to filter out people based on protected grounds.

Note that social media screening is not available in Québec. Québec’s privacy regime is more restrictive than the rest of Canada and most compliant providers exclude the province from this type of check.

What Employers Can Legally Review

Publicly available information, with limits. Public LinkedIn profiles, public posts, blogs, and portfolios are fair game in principle. “Public” is not a free pass, though. Employers still need to justify why the information is relevant to the role.

Job-relevant content. The narrower the focus, the safer the check. Examples of legitimately reviewable content include evidence of workplace misconduct, harassment or discriminatory behaviour, breaches of confidentiality from previous employers, public threats, or admitted illegal activity.

Information collected with consent. Best practice, and increasingly the expected standard, is to obtain written consent before screening a candidate’s social media. The consent should explain what is being reviewed and why. This is what separates a defensible process from an informal Google search by a hiring manager.

What Employers Cannot Do

Access private accounts. Asking for passwords, requesting that candidates open their accounts during an interview, or attempting to bypass privacy settings is not permitted. Full stop.

Use irrelevant personal information. Even when content is public, employers cannot use it if it has no bearing on the candidate’s ability to do the job. Lifestyle, relationships, religion, and political views fall into this category.

Discriminate on protected grounds. Social media often reveals information protected under federal and provincial human rights law: race, gender identity, sexual orientation, religion, disability, family status. Once a hiring manager has seen this information, it is very difficult to prove it did not influence the decision. That is precisely the risk.

Treat public access as consent. Information being available online does not give an employer unlimited rights to use it. The legal test is purpose, not accessibility.

Where Informal Screening Goes Wrong

Three patterns create most of the legal exposure:

The hiring manager screens from their personal account, which means they see all the protected information they shouldn’t and there is no audit trail of what was reviewed.

The screening is inconsistent. Some candidates get checked, others don’t, and the criteria shift based on who is doing the looking. This is how discrimination claims get built.

There is no documented consent and no record of what was flagged, what was ignored, and why. If a rejected candidate asks how the decision was made, the company has no answer.

Best Practices

Build a written social media screening policy and apply it to every candidate at the same stage of the hiring process. Define the risk categories you are screening for in advance. Get written consent before running a check. Use a third party to filter out protected information before it reaches the hiring manager. Keep a record of what was reviewed and what informed the decision.

How Credibled's Social Media Check Works

Credibled’s Social Media Check is built specifically to solve the compliance problem. Three steps:

The candidate consents to the screening before anything happens. AI scans public posts and content for predefined risk categories or custom keywords you set for the role. Only flagged content is included in the report. Every other post, photo, and comment stays out of the hiring manager’s view.

The risk categories include hate speech, insults and bullying, narcotics, drug-related and explicit imagery, terrorism and extremism, threats of violence, sexual impropriety, and toxic language. You can add custom keywords on top.

Reports are typically delivered in under an hour. Setup is mobile-first for both the employer dashboard and the candidate experience. Coverage is worldwide. Screening is not available for candidates in Québec.

The result: the hiring manager gets the job-relevant signal they were trying to find, without ever seeing the protected information that creates legal risk. That is the compliance wedge.

If you want to see what a compliant social media check looks like in practice, book a demo or reach out to [email protected].

Introduction

Most hiring managers treat the reference check as a formality. By the time you get to it, you have a preferred candidate, the offer letter is drafted, and the start date is on the calendar.

Then a reference says something that gives you pause.

What do you do with it? A single piece of negative feedback rarely tells the full story. But ignoring real signals is how bad hires happen. The skill is knowing the difference.

This is a guide for Canadian employers on how to read reference responses, when to dig deeper, and when a bad reference is serious enough to change your decision.

What actually counts as a bad reference

Not every less-than-glowing comment should worry you. A reference who says “she’s strong, though she sometimes needed prompting on deadlines” is being honest, not damaging. Hiring managers who expect perfect feedback often miss the more useful signal underneath. Real references give you context, not endorsements.

A reference becomes a genuine concern when you see one of three things.

Severity. Misconduct, dishonesty, repeated reliability failures, or behaviour that hurt the team. These are not minor.

Consistency. Two or three references raising the same theme. One person calling a candidate hard to work with is one data point. Three references using similar language is a pattern.

Contradiction. The reference describes a different role, scope, or tenure than the candidate did. This is the strongest signal of all and the easiest to verify.

If a response shows one of these, slow down.

The most common bad reference is silence

Before you worry about negative feedback, look at what you’re not getting.

The most common reference outcome at most companies is no response at all. References ghost. Forms get half-completed. Emails go to junk folders. Hiring teams accept candidates with two completed references instead of four and tell themselves the others were “unavailable.”

Silence is data. A reference who refuses to respond after multiple follow-ups, or who agrees to take a call and then stops returning messages, is signalling something. It does not always mean the candidate was a poor employee. It often means the relationship ended badly enough that the reference does not want to be on record.

This is why response rate matters as much as response content. A system that drives completion gives you more to work with. Credibled’s automated reminders run every 24 hours for up to 10 days, which closes the gap between requested and completed checks and exposes silence early enough to act on it.

Reading severity correctly

Some concerns are workplace friction. Others are disqualifying.

Concerns that usually rise to the level of changing a hiring decision:

Dishonesty about role, title, scope, or dates

Misconduct or workplace harassment

Theft, fraud, or breach of trust

A pattern of unreliability that affected team output

Inability to perform the core responsibilities of the role

Concerns that usually call for a follow-up conversation rather than a rejection:

Personality friction with one specific manager

A slow ramp in a previous role

Difficulty with a skill the candidate has since developed

One missed deadline cited without context

The first list deserves a second reference conversation and possibly a frank discussion with the candidate. The second list deserves curiosity, not alarm.

When contradiction shows up

If a candidate said they led a team of eight and the reference describes them as an individual contributor, you have a problem. Misrepresentation is the cleanest reason to withdraw a conditional offer. It is documented, defensible, and removes the awkward conversation about “fit.”

Most contradictions are not deliberate. Candidates round up their scope, inflate titles, or claim credit for projects they contributed to. A few are deliberate. Either way, the reference response is where the mismatch surfaces.

Structured digital reference checks make this easier to spot. When every reference answers the same questions in the same format, comparing their version to the candidate’s version is straightforward. Phone references rarely produce this clarity because the conversation drifts and the notes never line up.

When to withdraw an offer

A conditional offer can be rescinded when the conditions are not met. Reference outcomes are a standard condition.

Before you act, do three things.

Talk to the candidate. Share the concern without naming the reference. Ask for context. Sometimes the explanation reframes the issue. Sometimes the response confirms it.

Get a second data point. One reference saying something serious is worth investigating. Two references saying the same thing is worth acting on.

Document the decision. Write down what was said, what you verified, and why the offer is being withdrawn. This protects the company and forces clarity on whether the reasoning is sound.

If the issue is severity plus consistency, or any case of clear contradiction, withdrawing is usually the right call.

The legal landscape in Canada

In Canada, employers generally have latitude to share truthful reference information. Many limit their responses to employment confirmation because their internal policy is risk-averse, not because the law forbids more.

References should provide accurate information, avoid disclosing protected personal details, and stay clear of exaggerated or malicious statements. Employers requesting references should have candidate consent and follow applicable privacy requirements.

This is general information, not legal advice. Specific situations involving rescinded offers, defamation risk, or human rights considerations should be reviewed with employment counsel.

Why digital reference checks give better signals

Phone references produce polite, generic feedback because referees do not want to be cornered into criticism on a live call. The conversation moves quickly, the recruiter takes partial notes, and the same questions are not asked of every reference.

Written reference checks change the conditions.

References respond privately, on their own time. They give more thoughtful answers because they can think before they type. Every reference answers the same questions, which makes patterns and contradictions easier to see. Responses are documented, which protects everyone if a hiring decision is later questioned.

Credibled also flags duplicate screenings across recruiters at the same agency, so a candidate referred twice does not generate two paid checks.

A simple framework before you act

When a reference response gives you pause, ask:

Is the concern severe, or is it friction?

Is the concern repeated by other references, or is it isolated?

Does the concern contradict what the candidate told you?

Have I given the candidate a chance to respond?

If the answer to two or more is yes, the offer is not safe.

Conclusion

Reference checks are not a formality. They are the last structured chance to confirm what an interview suggested and to catch what an interview missed. A bad reference does not automatically end a candidacy. A serious, consistent, or contradicted one usually should.