Ontario's new AI disclosure rule for job postings: what HR needs to do now
As of Jan. 1, 2026, employers in Ontario with 25 or more employees must disclose whether artificial intelligence is used to screen, assess, or select applicants in any publicly advertised job posting. This applies on the day the posting goes live.
"Publicly advertised" means external postings open to the general public. It does not include general recruitment campaigns without a specific role, internal postings, or roles performed outside Ontario (including mixed-location roles where the extra-provincial work is not a continuation of work performed in Ontario).
What you must disclose
Every qualifying job ad must say whether AI is used to screen, assess, or select applicants. The law doesn't prescribe the exact wording or level of detail.
Since official guidance is still pending, aim for clear, honest language that reflects how your team actually uses AI in the hiring process. Document your rationale and keep it consistent across postings and vendors.
How broad is "AI," "screen," "assess," and "select" likely to be?
The regulations adopt an AI definition that aligns with the OECD: machine-based systems that infer from inputs to produce outputs (predictions, recommendations, or decisions) that can influence outcomes. That points to a broad scope.
In practice, plan to disclose if tools parse resumés, rank or score candidates, generate assessments, recommend shortlists, or otherwise influence progression-even if humans make the final call. Ordinary meanings likely apply to "screen," "assess," and "select," so don't overcomplicate it.
Examples you can adapt
- Minimal statement: "We use AI-enabled tools to assist with applicant screening during our hiring process."
- More detailed (still concise): "This posting uses AI-enabled tools to help screen and assess applications (e.g., resumé parsing and candidate ranking). Final decisions are made by our hiring team."
Choose an approach that fits your risk tolerance and communications style. More detail can increase transparency and candidate trust but may invite more questions-so prepare a consistent Q&A.
Update these documents and workflows
- Job posting templates: Add an AI disclosure field and approval step before publishing.
- Recruitment policy: Define when disclosure applies; outline wording options and review steps.
- Hiring training: Educate recruiters and managers on identifying AI use across tools and platforms.
- Third-party agreements: Require recruiters and job boards to flag AI features and include your disclosure language.
- AI policy: Capture approved tools, human oversight, bias testing, data handling, and disclosure obligations.
- Vendor inventory: Track where AI is embedded (ATS, assessment tools, sourcing platforms) and how it's configured.
Compliance risk and reputation risk
On conviction, corporations face fines up to $100,000 under the ESA, with higher penalties for repeat convictions. Even before penalties, public non-compliance can damage trust with candidates and hurt referral pipelines.
Practical best practices HR can implement this week
- Map your hiring flow: Identify where AI touches resumes, rankings, assessments, or recommendations-inside your ATS or via vendors.
- Pick your disclosure model: Minimal vs. detailed. Decide now and standardize across all postings.
- Create a posting checklist: Role location, employee count threshold, disclosure language, and reviewer sign-off.
- Centralize candidate FAQs: Have clear answers on where AI is used and how humans oversee decisions.
- Monitor updates: Track guidance from the Ministry of Labour and adjust templates promptly.
- Audit vendors quarterly: Features change; re-confirm whether AI is enabled and how it affects candidate progression.
- Log decisions: Keep records of why tools are or aren't considered "screening," "assessing," or "selecting."
Clarifying edge cases
- Internal postings: Not covered.
- General recruitment (no specific role): Not covered.
- Roles outside Ontario: Not covered, including mixed-location roles where the extra-provincial work is not a continuation of work performed in Ontario.
- Human in the loop: You still disclose if AI materially influences progression at any step.
Should you use blanket disclosure?
Given AI is increasingly embedded in common tools, many employers are opting for a simple, consistent disclosure on every public posting. It reduces the risk of missing a hidden AI feature and streamlines compliance.
If you go this route, align the statement with your AI policy and your actual tooling, and refresh it as your stack evolves.
Looking ahead
Expect more AI rules in employment at both provincial and federal levels. Use this requirement to tighten your AI governance across hiring: standardize oversight, vendor reviews, and communications.
If your team needs structured upskilling to put this into practice, see the AI Learning Path for HR Managers.
Disclaimer: This article is for general information only and is not legal advice. Consult counsel for guidance on your specific circumstances.
Your membership also unlocks:
