Since 1 January 2026, Ontario's Employment Standards Act, 2000 (ESA) has required employers with 25 or more employees to say so in publicly advertised job postings when they use artificial intelligence to screen, assess or select applicants. Ten months into Ontario AI disclosure in job postings, most employers have the basic idea. The questions that keep coming up are smaller and more practical: does our tool count, what exactly do we write, and where does it go?
This guide answers those questions, gives you three sample statements you can adapt today, and puts the rest of Ontario's 2026 posting rules in one table.
#What does Ontario's AI disclosure rule actually require?
Ontario's guide to the ESA is refreshingly short on this point. If you use AI to screen, assess or select applicants for a publicly advertised position, the posting has to say so. That's it. There's no mandated wording, no requirement to describe how the system works, and no form to file.
Ontario defines artificial intelligence broadly. The guide describes it as a system that "infers from the input it receives" to generate outputs such as predictions or decisions, and notes that whether something qualifies depends on the facts. In other words, the label on the box matters less than what the tool does with candidate information.
Two details catch people out:
- Third parties count. If a recruitment agency, a staffing partner or a software vendor uses AI to screen your applicants, the posting still needs the disclosure.
- The rule starts with the posting. It applies to postings made on or after 1 January 2026.
#Who has to include an AI disclosure in an Ontario job posting?
The 25-employee threshold is a headcount, not a full-time-equivalent calculation. Ontario counts individual employees on the day the posting goes up. Part-time and casual staff each count as one, and all of your Ontario locations are counted together. Probationary staff, term employees and people on leave or layoff count too.
Imagine a Mississauga contact centre with 14 full-time agents, 9 part-timers and 4 casual weekend staff. That's 27 people, so it's over the line, even though it may feel like a small team.
The rules don't apply to:
- employers with fewer than 25 employees on the posting date
- positions open only to existing employees (internal postings)
- general recruitment campaigns or "help wanted" signs that don't advertise a specific position
- positions where the work is done entirely outside Ontario
#Does your tool "screen, assess or select"? A quick checklist
HR Reporter notes that the Ministry hasn't defined "screen," "assess" or "select," which is why so many teams are unsure. Until there's more guidance, a conservative reading is the safe one. Run each tool in your hiring stack through these questions:
| Question | If yes… |
|---|---|
| Does it filter, rank or shortlist applicants (for example, by résumé keywords or knock-out answers)? | Likely screening. Disclose. |
| Does it score, grade or summarize applicant answers, tests or interviews? | Likely assessing. Disclose. |
| Does it recommend who to advance, reject or hire? | Likely selecting. Disclose. |
| Does a human rely on its output when deciding who moves forward? | Treat it as in scope. Disclose. |
| Does it only schedule interviews, send reminders or fix typos in your job ad? | Probably not in scope. Document your reasoning. |
If a tool sits in a grey zone, ask the vendor in writing what inputs it uses and what it outputs. If the output is anything about the candidate (a score, a rank, a flag, a summary), you're almost certainly in disclosure territory.
#Sample AI disclosure statements for Ontario job postings
Ontario only requires a general statement. Still, the content gap online is real: most guides tell you that you must disclose, not what to write. Here are three tiers. Pick the one that fits your brand and your tool, then have your counsel review it.
#Minimal (meets the general-statement standard)
This employer uses artificial intelligence to screen, assess or select applicants for this position.
Short, plain and close to the language of the rule. Good for job boards with tight character limits.
#Standard (adds the human role and a contact)
We use artificial intelligence (AI) to help screen and assess applications for this role, including scoring structured interview answers against the job's criteria. A member of our hiring team reviews the results and makes every hiring decision. Questions about how AI is used? Contact careers@yourcompany.ca.
This is the version we'd suggest for most employers. It's honest, it tells candidates what to expect, and it gives them somewhere to go.
#Detailed (for candidates who want the full picture)
How we use AI in hiring. For this position, applicants may complete a structured interview by voice, on camera; it is recorded so our hiring team can review it, and recordings are deleted after [180 days]. An AI tool scores each answer against the same job-related rubric for every candidate and highlights quotes from the answer that support each score. It does not analyze faces, voice tone or emotion. A recruiter reviews every score and the supporting evidence, and a person makes all decisions about who advances. If you need an accommodation (such as typed answers or an interview without video) or would prefer a different interview format, contact careers@yourcompany.ca. Learn more in our candidate privacy notice: [link].
The detailed version is longer than Ontario requires, but it can reduce candidate anxiety and save your recruiters from answering the same email 40 times a week. Only describe what your tool genuinely does.
#How common is AI disclosure in Ontario job postings now?
Very common, and growing fast. Indeed Hiring Lab found that the share of Ontario job postings mentioning AI in the hiring process rose from 9% in October 2025 to 28% in May 2026. The rest of Canada rose from 4% to 12% over the same period, which Indeed described as a spillover from Ontario's policy. AI mentions rose in all 48 occupations Indeed tracked in Ontario.
Two takeaways. First, candidates in Ontario are getting used to seeing these statements, so a clear one won't scare anyone off. Second, national employers are increasingly using one disclosure across provinces rather than maintaining Ontario-only versions. That's often simpler, as long as the statement is accurate everywhere.
#What are the other Ontario job posting requirements for 2026?
The AI disclosure arrived alongside several other changes. Here's the full set for publicly advertised postings by employers with 25+ employees:
| Requirement | What it means in practice |
|---|---|
| AI disclosure | State that AI is used to screen, assess or select applicants, if it is. |
| Pay transparency | Include the expected pay or a pay range. A range can't span more than $50,000 a year. Not required if the pay (or top of the range) is above $200,000 a year. |
| Vacancy statement | Say whether the posting is for an existing vacancy or not (for example, a talent pool). |
| No "Canadian experience" | Postings and application forms can't require Canadian work experience. Licensing and registration requirements aren't caught by this. |
| 45-day reply | If you interview someone, tell them whether a hiring decision has been made within 45 days of the interview (or the last interview). Email, text, phone or a portal update with a link all work. |
| Record keeping | Keep copies of postings, application forms and the information given to interviewed applicants for three years. |
The 45-day rule is the one most likely to trip up high-volume teams. If a Mississauga contact centre interviews 400 people for a seasonal campaign, every one of them needs an update within the window. Automated status messages through your applicant tracking system are your friend here.
#How MSZ Hire fits into Ontario's rules
If you use AI interviews, you're in scope, full stop. MSZ Hire is built to make the disclosure part easy rather than to argue it away: it generates jurisdiction-aware notices and consent for Ontario (as well as Quebec, NYC, Illinois, Colorado and California), so the candidate sees the right language before the interview starts. Interviews are structured and spoken, with live captions and accommodations on request, and every score comes with quoted evidence a recruiter can check. A person makes every decision.
If you hire in Quebec too, the rules there work differently. Our guide to Quebec Law 25 and AI interviews explains when section 12.1 applies, and our plain-language map of PIPEDA and Bill C-36 covers the privacy side across Canada.
#Your Ontario AI disclosure checklist
- List every tool that touches applicants, including agency and vendor tools.
- Run each one through the screen/assess/select checklist above.
- Pick a disclosure tier and get it reviewed.
- Add it to every public posting template, alongside the pay range and vacancy statement.
- Set up 45-day status updates for interviewed candidates.
- Archive postings and candidate notices for three years.
Curious what a disclosed, structured AI interview feels like from the other side? Try the candidate experience yourself and see the notice, the questions and the format a candidate would see. When you're ready to bring structured, evidence-based AI interviews to your Ontario hiring, with the notices built in and a human making every call, MSZ Hire is here to help.
This article is general information, not legal advice. Talk to an Ontario employment lawyer about your specific situation.