Ontario's AI Hiring Disclosure Rule Took Effect This Year. Most Job Postings Still Don't Mention It.

Since January 1, 2026, Ontario employers with 25 or more employees have been required to say, directly on any public job posting, whether the process behind it uses artificial intelligence to screen, assess, or select candidates. The requirement is part of the Working for Workers Four Act, passed in 2024 and phased into the Employment Standards Act, and it arrived alongside a more widely publicized change requiring those same postings to include an actual wage or salary range rather than the vague "competitive salary" language that has served as a placeholder for years. The AI disclosure line got far less attention when the bill passed. It is now, quietly, an active legal obligation, and a large share of the job postings currently running in the province do not include it.

The reason it is easy to miss is that "using AI to screen candidates" describes a lot of hiring technology most businesses never labeled that way. A company that adopted an applicant-tracking system with a built-in resume-ranking feature, a video-interview platform that scores candidates on tone or word choice, or a chatbot that conducts an initial screening conversation before a person ever reviews the application is, for purposes of this law, using AI to screen or assess. Many of these features arrived as quiet additions to tools a business had already been using for years — the same pattern regulators have been pursuing under AI-washing enforcement, except here the disclosure duty runs toward the employer, not the vendor. A hiring manager who has never thought of the ATS as "an AI tool" can be running a process the law now requires the business to disclose.

The disclosure itself is not onerous. A plain sentence on the posting, stating that AI is used in the screening or selection process and what it is used for, satisfies the core requirement. What trips businesses up is the record-keeping attached to it: Ontario now requires job postings and the communications tied to them to be retained for three years, which means a business needs an actual process for capturing that, not a folder of ads that gets cleared out once a role is filled. For a business that fills a handful of roles a year, that is a minor administrative habit to build. For one running frequent postings through a recruiter or a job board, it is worth confirming who is actually responsible for keeping the record, because "the recruiter has it somewhere" is not the same as the employer holding it.

This requirement does not stand alone. Ontario employers with 25 or more employees have had to maintain a written electronic monitoring policy since 2022, disclosing whether and how they track staff — and that policy was written for an earlier generation of monitoring tools: login times, keystrokes, location data. A number of businesses have since layered AI on top of that monitoring, through productivity-scoring software, AI-summarized activity reports, or tools that flag "disengagement" from patterns in an employee's digital activity, without revisiting the policy that is supposed to disclose it. The guidance around that requirement treats AI-based monitoring as squarely within what the policy needs to describe. A business that added an AI feature to an existing monitoring tool, without updating the document that discloses it, has the same kind of quiet gap the hiring-disclosure rule is now closing on the recruitment side.

None of this is being enforced from far away or on a distant timeline. Employment standards officers in Ontario conduct proactive inspections, and both of these obligations sit inside the Employment Standards Act — the same statute that already covers wages, hours, and termination pay, not a separate AI-specific regime with a soft rollout period. A Mississauga business that has spent the past year keeping an eye on the EU AI Act, or a Canadian federal AI statute that never actually passed, can end up missing the one that is already enforceable against it, this year, in its own province.

The practical fix is a short internal audit, not a legal overhaul. List every tool involved in hiring and staff monitoring, and ask plainly whether any of it uses AI to score, rank, flag, or summarize a person — an applicant or an employee. Where the answer is yes on the hiring side, add the disclosure line to every external posting and put an actual retention habit in place for those postings and the communications tied to them. Where the answer is yes on the monitoring side, update the written electronic monitoring policy to say so specifically, rather than relying on language drafted before the AI feature existed. Neither task requires outside counsel to complete, though it is worth having someone confirm the final wording before it goes live.

The businesses that get caught out here will not be the ones that adopted AI hiring or monitoring tools aggressively. They will be the ones that adopted them quietly, a feature at a time, inside software they had already been using for years — and never went back to check whether the disclosure obligations that came with those features had already taken effect.