AI at Work in Ontario: Your Rights When Algorithms Hire, Monitor, or Fire You

AI has quietly moved into every stage of the employment relationship in Ontario: resume screeners that rank applicants before a human sees them, productivity software that scores your keystrokes and idle time, and analytics tools that flag "underperformers" for restructuring lists. Employer-side law firms are busy advising companies on how to roll these tools out. This guide covers the other side: what you, the employee, are entitled to know, and what happens to your rights when the decision-maker is a piece of software.

What must a job posting tell you about AI screening?

The Working for Workers Four Act, 2024 (Bill 149) amended the Employment Standards Act to add new rules for publicly advertised job postings. Effective January 1, 2026, an employer that uses artificial intelligence to screen, assess, or select applicants for a publicly advertised position must say so in the posting. The requirement applies to employers with 25 or more employees, and it arrived alongside related posting rules, including mandatory expected pay ranges and a ban on requiring "Canadian experience."

Two things to understand about this disclosure right. First, it is a transparency rule, not a prohibition: employers can still use AI to rank and reject applicants, they just have to tell you it is happening. Second, the disclosure sits in the posting itself, so read postings carefully. If you see an AI disclosure, assume your resume needs to survive software before a human reads it, and assume the employer has data about how you were scored. That data matters later if you suspect the screening was discriminatory.

Can your employer legally monitor you with AI?

Mostly yes, but you are entitled to know how. Since the Working for Workers Act, 2022 (Bill 88), s. 41.1.1 of the ESA has required every Ontario employer with 25 or more employees to have a written electronic monitoring policy. The policy must state whether the employer electronically monitors employees, and if it does, describe how and in what circumstances it monitors and the purposes for which the information may be used. You are entitled to a copy of the policy within 30 days of being hired, and to an updated copy when it changes.

The honest caveat: the policy requirement does not cap what monitoring an employer can do, and it does not create a standalone privacy right for most provincially regulated Ontario employees. But it is far from useless. The policy tells you whether your keystrokes, GPS location, emails, or screen time are being scored, which is exactly the data that AI performance tools feed on. If your employer disciplines or fires you based on surveillance data it never disclosed, that gap becomes leverage. We cover the broader surveillance picture in our guide to employer monitoring and recording at work in Ontario.

What if AI screening discriminates against you?

This is the sharpest legal edge in workplace AI. Section 5 of Ontario's Human Rights Code guarantees equal treatment in employment without discrimination based on grounds like age, disability, race, sex, and family status. The Code catches adverse effect discrimination: a rule or tool that looks neutral but disproportionately screens out a protected group. Intent is irrelevant. An employer cannot answer a discrimination claim by saying the algorithm did it, and it cannot outsource liability to the software vendor it chose.

• A resume screener trained on the employer's past hires can learn to prefer the demographics of those hires, quietly filtering out older workers or racialized candidates.
• Tools that penalize employment gaps can disproportionately screen out people who took disability leave, parental leave, or caregiving time, all connected to protected grounds.
• Video-interview analysis that scores speech patterns or facial expressions can penalize candidates with accents, autism, or speech-related disabilities.
• Productivity-monitoring scores that ignore accommodations (modified hours, breaks for medical needs) can flag accommodated employees as underperformers, turning an accommodation into a termination risk.

If you believe an AI tool screened you out or scored you down because of a protected characteristic, you can file an application with the Human Rights Tribunal of Ontario, generally within one year of the event. Compensation can include lost income and damages for injury to dignity. Our guides on workplace discrimination in Ontario and human rights damages in employment cases walk through how those claims work and what they pay.

Can an algorithm actually fire you?

Legally, no algorithm fires anyone. Your employer fires you, and your employer answers for it, whatever software sat behind the decision. That framing matters because employees sometimes assume a data-driven termination is somehow more legitimate or harder to challenge. It is not. If you are terminated without cause in Ontario, you are owed your ESA minimums plus, in most cases, common law reasonable notice, assessed on the same factors courts have used for decades (the Bardal factors): your age, length of service, the character of your position, and the availability of similar employment. Nothing in that framework asks who, or what, recommended the termination.

Where AI shows up most dangerously is in alleged cause. Some employers point to algorithmic productivity scores, idle-time metrics, or automated performance rankings to justify a for-cause termination that pays nothing. Cause is the capital punishment of employment law and the bar is very high: the employer must prove misconduct or incompetence serious enough to destroy the employment relationship, usually after warnings and a real chance to improve. A dashboard score, generated by a tool the employee never saw, measuring proxies the employee never agreed to, will rarely clear that bar. And where the process was cold, automated, misleading, or factually wrong, the manner of dismissal itself can attract additional moral damages under the principles from Honda v. Keays. If any of this resembles your termination, start with our guide to wrongful dismissal in Ontario.

What are you entitled to know, and what should you ask for?

Employer-side compliance guides tell companies the minimum they must disclose. Flip that around and you get your entitlement checklist:

• Job postings: if the employer has 25 or more employees, the public posting must disclose AI screening. No disclosure plus evidence of AI screening is an ESA compliance problem you can raise with the Ministry of Labour.
• Monitoring policy: you are entitled to the written electronic monitoring policy within 30 days of hire. Ask for it in writing if you never received it; the request itself is protected activity, and reprisal for asserting ESA rights is prohibited.
• How a decision was made: when discipline or termination lands, ask in writing what information the decision was based on and whether a human being reviewed it. Employers are not always required to answer, but the answer (or the silence) becomes evidence.
• Your own data: keep contemporaneous copies of performance dashboards, productivity scores, rankings, and any emails referencing them. Do this before you lose system access; termination usually cuts access the same day.
• Your severance: an AI-assisted termination changes nothing about your entitlements. Two weeks per year of service is a myth in both directions; common law notice often runs months per year for many employees.

What should you do if AI played a role in your termination?

1. 1.Do not sign the termination letter, severance offer, or release on the spot. Deadlines in offers are almost always negotiable, and signing releases your claims.
2. 2.Ask in writing what the termination decision was based on, what data or tools were used, and whether a human reviewed the outcome before it was final.
3. 3.Gather your evidence: the original job posting, the electronic monitoring policy, performance dashboards and scores, your reviews, and any emails about metrics or rankings.
4. 4.Map your deadlines: HRTO applications generally within one year for discrimination, and wrongful dismissal claims generally within two years.
5. 5.Get the offer professionally reviewed before you respond. Most severance offers open well below what the law requires, and AI-justified terminations often come with weak cause allegations that collapse under scrutiny.

The technology is new; your rights are not. Whether a manager or a model made the call, Ontario law measures your exit package the same way, and most first offers fall short. Start with our plain-language guide to severance pay in Ontario, and if you have an offer in hand, our free severance review will tell you within a day whether the algorithm's employer is trying to underpay you.

About the Author

Daniel Carter

Legal Writer, Mirza Law

Daniel Carter is a legal writer at Mirza Law in Toronto. He writes about layoffs, employment contracts, and the steps to take before you sign anything from your employer.

See all articles