The federal privacy law most often invoked when the subject of drug and alcohol testing comes up usually doesn’t apply to the construction employer asking the question at all. The boundary that actually governs testing on a Canadian job site runs through human rights law and how carefully a safety policy is written — not through the statute most people reach for first.
Key takeaways
PIPEDA gets raised constantly in discussions of workplace drug and alcohol testing, and for most construction employers it’s the wrong statute to be checking. The Office of the Privacy Commissioner’s own guidance on workplace privacy is explicit that its detailed framework covers “federally-regulated employers” — banks, telecommunications carriers, and other federal works and undertakings — and notes separately that “several provinces have privacy legislation applying to employee information,” without providing detailed guidance for those provincially-regulated employers, which is what the overwhelming majority of construction firms are. That doesn’t mean privacy considerations vanish for a provincially-regulated contractor — it means the specific federal statute most often quoted isn’t the one doing the governing, and a firm relying on PIPEDA compliance as its defence for a testing policy is relying on the wrong document.
Where provincial privacy legislation or common-law privacy principles do apply, the OPC’s own balancing framework is worth understanding regardless of which specific statute governs, because it reflects the reasoning courts and tribunals tend to apply anyway: how sensitive is the personal information being collected, is the employer’s purpose a genuine business need rather than a convenience, is there a less invasive way to achieve the same protective goal at comparable cost, and is the resulting loss of privacy proportional to the benefit the employer is actually getting. Applied to drug and alcohol testing, that framework tends to favour testing tied to a specific, demonstrable safety risk — a genuinely safety-sensitive role, a post-incident investigation, reasonable cause based on observed behaviour — over blanket or purely random testing applied without that connection.
Ontario’s Human Rights Code, and the Ontario Human Rights Commission’s own published policy on discrimination based on mental health disabilities and addictions, frame substance addiction as a disability subject to the same duty to accommodate as any other disability, up to the point of undue hardship for the employer. That has a direct, practical consequence for testing policy: a positive result can’t automatically translate into discipline or termination without first considering whether the underlying issue is addiction, and if so, what accommodation looks like before more serious steps are taken. A policy built purely around “test positive, get disciplined” skips a legal step that exists specifically because addiction is treated as a health condition, not a character flaw, under the Code.
CCOHS frames the core of a workable policy as something decided before any testing happens, not improvised in response to an incident: a clear statement of the circumstances under which testing will be conducted, how it will be conducted — oral, breath, urine, or some combination — the scope of substances being tested for, and the criteria used to interpret the results. The value of deciding all of that in advance isn’t just legal tidiness. A policy drafted case by case, in the moment, is far more likely to be applied inconsistently between workers — which is itself a discrimination risk independent of whether the testing itself was justified.
Put together, the practical shape of a defensible policy is narrower than most firms assume: identify which roles are genuinely safety-sensitive rather than labelling every position that way, tie testing triggers to demonstrable safety risk rather than convenience, build in an accommodation step for a positive result before jumping to discipline, and write every element of the method and interpretation down before the first test is ever administered. None of that is a decision a scheduling or HR tool should make on its own — the same rule that applies everywhere else in this hub applies here: a system can flag when a policy trigger has been met, but a person has to decide what happens next, and sign off on it. Firms building out a broader safety programme will find the return-to-work and injury-cost pieces in this hub’s safety cluster cover the accommodation side of this problem in more depth.
Related reading: the accommodation process a positive test or an injury both feed into and how an unmanaged safety incident compounds beyond the incident itself.
PIPEDA generally doesn’t govern the question either way for most construction employers, because it applies to federally-regulated organizations and most construction firms are provincially regulated. Provincial privacy principles and human rights law are the frameworks that actually apply.
Blanket or purely random testing without a demonstrable, specific safety justification is the hardest form of testing to defend under the proportionality principles privacy law applies, and it doesn’t address the separate duty to accommodate addiction as a disability. Testing tied to a genuinely safety-sensitive role, post-incident investigation, or observed reasonable cause is on firmer ground.
Per CCOHS guidance, a policy should set out in advance the circumstances that trigger testing, the testing method (oral, breath, urine, or a combination), the substances covered, and how results will be interpreted — decided before any test, not case by case.
Yes. The Ontario Human Rights Commission treats addiction as a disability under the Human Rights Code, which means an employer has a duty to accommodate it up to the point of undue hardship, including before treating a positive test result as grounds for discipline.
A 30-minute call is enough to tell you whether AI pays for itself here.
