Anonymised, illustrative composite. A Barrie brokerage's social-media manager built a listing ad audience meant to reach “serious buyers” by excluding renters and anyone flagged as receiving government assistance. It read, on the ad platform's own dashboard, as a targeting optimization. Under Ontario's Human Rights Code, it read as something else.
At a glance
The brokerage's social-media manager, building an ad for a move-up detached listing, wanted to avoid wasting spend on people unlikely to qualify for the price point. Working inside a social platform's ad-targeting tool, they built an audience by excluding several interest and behavioural categories the platform offered, including segments the platform itself labelled around rental housing search activity and public-assistance-adjacent interests, reasoning that anyone in those categories was unlikely to be shopping in this price bracket anyway.
The ad ran for two weeks. No individual complained of being personally excluded — the platform's targeting tools don't work that way, and no one sees the audience definition behind an ad they never received. The complaint instead came from a housing-advocacy organization that audits real estate ad targeting practices and flagged the campaign's audience settings, visible through the platform's ad-transparency library.
Ontario's Human Rights Code protects a specific list of grounds in housing, and receipt of public assistance is on that list alongside “race, colour or ethnic background; religious beliefs or practices;… family status; marital status… disability; sexual orientation; age… receipt of public assistance.” The Commission's own policy gives an advertising example directly on point: a rental ad reading “suits a working person,” implying that people who receive social assistance… are not welcome or need not apply is treated as an example of discriminatory advertising.
Excluding an ad audience by a category correlated with public assistance is not the same act as printing “suits a working person” on a flyer, but it reaches toward the same substance the Commission's example targets: steering who does and does not see a housing opportunity based on a protected characteristic, rather than on ability to purchase at the listed price. The mechanism is automated and invisible to the excluded person, but the effect — some people never see the ad at all, sorted by a protected ground — is the thing the Code's housing provisions are aimed at.
This sits alongside, not inside, RECO's own rules: Bulletin 1.1 names discrimination under Ontario's Human Rights Code as conduct that breaches an agent's professional-conduct obligations, alongside misrepresentation and fraud — so a Human Rights Code problem in a brokerage's advertising is also, independently, a RECO professional-conduct problem.
The campaign ran for two weeks at a modest daily budget, targeting a metro-area audience narrowed by the excluded categories in question. No figures exist for how many people the exclusions actually removed from the eligible audience — that calculation lives entirely inside the ad platform's own systems and was not something the brokerage could reconstruct after the fact.
The listing itself sold within the two-week window regardless, to a buyer who found it through a portal search rather than the social ad — a detail that did not change the underlying targeting question, since the Code's concern is with who was excluded from seeing the opportunity, not with whether the exclusion changed the sale's outcome.
The advocacy group's complaint did not need to show that any specific person had been denied housing. Ontario's housing-discrimination framework is aimed at the advertising and application process itself — the Commission's protected-grounds list and its advertising example both frame the harm as occurring at the point information about housing is withheld from a protected group, not only once a specific transaction is denied.
The determining fact, once the platform's ad-transparency data was reviewed, was simply that the excluded categories tracked closely enough to a protected ground — receipt of public assistance — that the targeting choice could not be explained purely as a proxy for purchasing power without also functioning as a proxy for that protected characteristic.
The brokerage removed the exclusion categories from its ad-templates library and adopted a rule limiting audience narrowing to purely geographic and price-band criteria going forward, with any behavioural or interest-based exclusion requiring sign-off from the broker of record before a campaign launches.
The matter was resolved through the brokerage's corrective changes rather than a formal Human Rights Tribunal application, but it prompted a wider review of every active social campaign's audience settings across the brokerage's other listings.
For the print-advertising side of the same Code of Ethics obligation, see a farming flyer that named a neighbour’s price, and on the underlying professional-conduct standard, the Code of Ethics glossary entry.
The tell was treating an ad platform's targeting categories as neutral cost-efficiency tools rather than as proxies that can map onto protected grounds. A category labelled around rental-housing search behaviour or public-assistance interests is not labelled “receipt of public assistance” on the dashboard — but the Code's own advertising example is precisely about language and choices that carry that implication without saying it outright.
A short call is enough to see how AI-assisted campaign review flags audience settings before a launch, not after a complaint.
