Ontario's New AI Disclosure Rule Changes Hiring for Canadian Employers
Starting January 1, 2026, Ontario employers with 25 or more employees must disclose when they use artificial intelligence to screen, assess or select job applicants. The requirement comes from the Working for Workers Four Act, 2024, which amended Ontario's Employment Standards Act.
The rule doesn't ban AI in hiring. It requires employers to tell candidates that AI is being used. Ontario is currently the only Canadian province with legislation explicitly mandating this disclosure.
The Regulatory Patchwork in North America
No single federal law in Canada or the United States governs AI use in hiring. Instead, employers face a fragmented system of federal, provincial and state rules.
In Canada, whether federal or provincial law applies depends on the employer's jurisdiction. Federally regulated employers - banks, telecommunications companies, airports - must comply with federal human rights, pay equity and accessibility laws alongside privacy rules. Provincial employers follow their province's employment standards and human rights codes.
The United States has no comprehensive federal AI law for employment. On March 20, 2026, the White House released a "National AI Legislative Framework" with recommendations for Congress, but it imposes no new employer obligations. California, Illinois, New York City and Texas have enacted their own state and local rules regulating AI in hiring, performance management and other employment decisions.
What U.S. Laws Currently Require
Existing antidiscrimination laws apply fully to AI-driven hiring decisions. Title VII of the Civil Rights Act, the Americans with Disabilities Act and state civil rights statutes all cover AI systems.
Several states have amended their antidiscrimination laws to explicitly prohibit discriminatory AI use. California's Consumer Privacy Act now includes regulations on automated decision-making technology. Under these rules, employers must conduct risk assessments before using AI to make significant employment decisions, such as hiring or denial of employment.
New York City requires bias audits for certain AI employment-decision tools. Other jurisdictions strongly encourage them through enforcement patterns. Class-action lawsuits against employers challenging AI hiring practices have increased, often citing existing fair credit reporting and antidiscrimination laws.
Privacy and Data Obligations
AI hiring tools process sensitive personal data. California's updated CCPA regulations require employers to provide notice before using AI to make employment decisions and to grant applicants access rights to that information.
Employers must determine whether their AI use qualifies as regulated automated decision-making technology. If it does, they must give notice before using the system and allow access to how decisions were made. Opt-out rights generally don't apply to hiring because employers cannot legally discriminate in employment decisions.
Discrimination and Bias: The Central Risk
Regulators across jurisdictions focus heavily on preventing discrimination and bias in AI hiring systems. Human rights laws prohibit AI tools from discriminating on grounds like race, sex, gender, ethnicity, creed or sexual orientation.
Accessibility laws require that AI-enabled hiring processes - including AI-led interviews - remain accessible to people with disabilities. Employers must ensure their systems work for applicants with hearing or visual impairments.
Documented bias testing and risk assessments are critical. Even where not legally mandated, they provide evidence of responsible practices if hiring decisions are later challenged as discriminatory.
Canada's Public Sector AI Oversight
Ontario introduced additional requirements for public sector employers through Bill 194, Strengthening Cyber Security and Building Trust in the Public Sector Act, 2024. The law sets expectations for ministries, universities, school boards and similar institutions around AI governance, transparency and technical standards.
Five Steps to Reduce Legal Risk
Inventory your AI use. Document which AI tools your organization uses across recruiting, performance management, employee monitoring and workforce planning. Classify each use case by legal risk and jurisdictional exposure.
Assess before deployment. Evaluate potential bias, data quality, explainability and whether human review will be meaningful or merely formal. Regulators expect genuine human oversight, not a checkbox exercise.
Keep humans in control. AI should inform decisions, not replace them. Human decision-makers must retain authority to override AI outputs and understand the tool's capabilities and limitations. Document escalation and review processes.
Manage vendors carefully. Employers remain responsible for hiring outcomes even when third parties provide AI tools. Contracts should address bias testing, audit rights, data protection, liability allocation and vendor cooperation in regulatory inquiries. Vendor assurances alone won't satisfy regulators.
Align policies and train staff. Update employee handbooks, privacy notices, candidate disclosures and internal AI policies to match actual practices. Train HR professionals, legal teams and business users on how the tools work and when to escalate concerns.
Compliance Is Ongoing
AI compliance is not a one-time project. Models drift over time. Laws change. A system compliant at launch may violate regulations six months later.
Employers should monitor for model drift, emerging bias and changes in applicable law. Continuous reassessment keeps hiring practices aligned with evolving legal requirements.
For HR professionals seeking deeper knowledge, resources on AI for Human Resources and an AI Learning Path for CHROs can help build organizational strategy around responsible AI deployment.
Your membership also unlocks:
