This article affirms that public policy must protect equal dignity, civil rights, and fair treatment under the law for every person in Canada — regardless of race, ethnicity, gender, sexual orientation, disability, religion, national origin, socioeconomic status, or any other characteristic protected under the Canadian Charter of Rights and Freedoms and federal and provincial human rights legislation.
Equality is not sameness of treatment. Delivering equal outcomes often requires unequal inputs: more support for those who face greater barriers. Policies that treat everyone identically while ignoring structural disadvantage do not produce equality — they reproduce it.
The digital dimension of equality is now central. Algorithmic systems used in public services — benefit eligibility, risk scoring, child welfare screening, border processing — have documented patterns of discriminatory output. Data about people is collected, retained, and used in ways that affect their rights without their knowledge or meaningful consent. Equality in the present century requires addressing automated and data-driven discrimination with the same seriousness as direct discrimination.
1. Equal dignity precedes policy. Rights are not granted by government — they are recognised. Public institutions are obligated to protect rights, not to decide which groups deserve protection.
2. Structural disadvantage is real and measurable. The causes of inequality are not limited to individual prejudice. Historic dispossession, systemic policy failures, and institutional design create disadvantages that persist across generations. Acknowledging these causes is a precondition for addressing them.
3. Rights review is built in, not added later. Equality considerations must be integrated into policy design from the beginning — not reviewed as an afterthought after implementation decisions have already been made.
4. Automated systems are subject to equality obligations. Algorithmic tools in public decision-making are not neutral. Discriminatory outputs are discriminatory outcomes regardless of whether a human or a system produced them. The same standards apply.
5. Data rights are civil rights. People have the right to know what data about them is collected, how it is used, and to have inaccurate data corrected. Personal data held by public institutions is not an administrative asset — it is information about people with rights over it.
6. Access to remedy matters as much as the right. A right without an accessible enforcement pathway is not a functional right. Complaint and remedy mechanisms must be adequately resourced and genuinely reachable by the people most likely to need them.
Mandatory rights-impact assessments All major federal programs, legislation, and regulatory changes must undergo a rights-impact assessment before implementation. Assessments must use a standardised methodology, be published in full, and be reviewed by an independent body. Identified rights risks must be addressed before implementation or publicly justified in writing if overridden.
Algorithmic accountability for public systems Federal agencies using algorithmic or automated decision-support tools in matters affecting individual rights must:
- Register the system in a public algorithmic register with decision criteria, training data sources, and validation methodology disclosed.
- Conduct and publish annual bias audits against protected characteristics under the Canadian Human Rights Act.
- Provide a free human review pathway for any algorithmically assisted decision affecting an individual, available on request.
- Notify affected individuals when an algorithmic system contributed to a decision about them.
Disaggregated data collection and publication Federal data programs collect and publish disaggregated data by race, gender, disability status, and other relevant characteristics, consistent with privacy protections and Indigenous data sovereignty principles. Policy evaluation cannot measure equality outcomes without data that reveals differential impacts. Disaggregated data is published in machine-readable formats.
Expanded accessibility standards Federal services, digital platforms receiving federal funding, and public procurement contracts must comply with WCAG 2.1 AA accessibility standards as a verified minimum. Compliance is independently audited, not self-reported. Non-compliance is a reportable breach with a remediation timeline, not a deferred aspiration.
Human rights complaints infrastructure The Canadian Human Rights Commission receives funding sufficient to resolve complaints within published timelines. Median complaint resolution within 12 months is a mandatory reporting metric. Systemic investigation authority is expanded so patterns can be addressed without requiring individual complainants to carry the full burden of initiating proceedings.
- Rights-impact assessments published for 100% of major federal policy packages within two years of this article's adoption.
- Federal algorithmic register operational within 18 months, with annual bias audit results published for all registered systems.
- Human rights complaint median resolution time falls to 12 months or fewer within three years.
- Disaggregated outcome data published for major federal transfer programs within two years.
- Independent accessibility audits of federal digital services show compliance above 90% within three years.
- Service barrier metrics for disabled and marginalised residents show measurable improvement across consecutive reporting cycles.