L'algorithme YieldStar de RealPage aurait permis à des propriétaires canadiens de coordonner des hausses de loyer
An algorithm that pools confidential data from competing landlords to generate coordinated pricing recommendations is the subject of antitrust investigations in both the US and Canada. The US DOJ reached a settlement with RealPage in November 2025, and Canada's Competition Bureau opened its own investigation in September 2024. RealPage has stated the software affects less than 1% of the Canadian rental market.
Récit
Since at least 2017, Canadian landlords have used RealPage’s YieldStar algorithm — a revenue management system that aggregates confidential rental data from competing property managers and generates coordinated rent price recommendations. An investigation by The Breach in September 2024 revealed that at least 13 Canadian companies with $5+ billion in revenue were using the software.
YieldStar collects competitively sensitive data that landlords would not normally share — current and historical rental prices, occupancy rates, signed leases, renewal offers, and future occupancy projections — and runs it through an algorithm that generates rent recommendations. RealPage marketed the software as delivering 3-7% outperformance versus market rates. For 22 John Street in Toronto’s Weston neighbourhood, YieldStar recommended annual increases ranging from 7% to 54%. The landlord, Dream Unlimited, implemented increases at the lower end — 9-10% annually — still far exceeding Ontario’s 2.5% rent control guideline, under a November 2018 policy that exempted newer units from rent control.
Canada’s Competition Bureau opened an investigation in September 2024. In December 2024, a proposed class action was filed on behalf of affected tenants (lead plaintiff: Cynthia Black), naming RealPage and 14 Canadian landlords including Dream Unlimited, GWL Realty Advisors, CAPREIT, Tricon Residential, and Choice Properties REIT. Federal Minister François-Philippe Champagne publicly called the situation “completely unacceptable” and urged the Competition Commissioner to act.
RealPage has stated that its software affects less than 1% of the Canadian rental market. Several landlords voluntarily discontinued YieldStar after media scrutiny: GWL Realty Advisors terminated use after an internal review in October 2024, and Dream Unlimited and Tricon Residential followed. In the United States, the DOJ reached a settlement with RealPage in November 2025, effectively banning its core business model of pooling nonpublic landlord data for rent recommendations — the outcome of the US case may be relevant to the Canadian investigation.
Préjudices
RealPage's YieldStar algorithm aggregated confidential, competitively sensitive rental data from competing landlords and allegedly generated coordinated rent recommendations, with documented suggestions ranging from 7% to 54% annual increases for a Toronto building — far exceeding Ontario's 2.5% rent control guideline for controlled units. Canadian renters experienced rent increases of 9-10% annually at buildings using the software.
Low-income tenants and renters in non-rent-controlled units faced unaffordable rent increases driven by algorithmic pricing recommendations, with documented cases of tenants working multiple jobs to afford rent at buildings using YieldStar.
Populations touchées
- Canadian renters in buildings managed by landlords using YieldStar, particularly in Ontario
- low-income tenants in non-rent-controlled units
Entités impliquées
A développé et exploité l'algorithme de gestion des revenus YieldStar ; a agrégé des données locatives confidentielles de propriétaires canadiens concurrents pour générer des recommandations de loyer coordonnées ; a affirmé que le logiciel touche moins de 1 % du marché locatif canadien
A ouvert une enquête sur la fixation algorithmique des prix des loyers en septembre 2024 ; a déclaré que la protection de la concurrence dans le secteur immobilier est une priorité
Systèmes d'IA impliqués
Algorithme de gestion des revenus qui regroupait des données confidentielles sur les prix, l'occupation et les baux de propriétaires concurrents pour générer des recommandations de loyer coordonnées ; suggestions documentées de hausses annuelles de 7 à 54 % pour des propriétés torontoises
Réponses et résultats
A ouvert une enquête sur l'utilisation par des propriétaires canadiens de logiciels de tarification locative algorithmique pour fixation potentielle des prix
Contexte du système d'IA
YieldStar is RealPage's revenue management algorithm that collects confidential rental data — including pricing, occupancy rates, signed leases, and renewal offers — from participating landlords and generates rent price recommendations. RealPage marketed it as delivering 3-7% outperformance versus market rates.
Mesures préventives
- Prohibit algorithmic systems that aggregate competitively sensitive pricing data from competing firms to generate coordinated price recommendations
- Require transparency and regulatory approval for algorithmic pricing systems used in essential markets such as housing
- Mandate that competition authorities develop specific enforcement frameworks for algorithmic collusion
- Ensure rent control legislation covers algorithmic pricing tools that circumvent provincial guidelines
Fiches connexes
Taxonomie
Sources
- Competition Bureau investigating price-fixing by Canadian landlords
- Canadian mega-landlord's AI pricing scheme hikes rents
- Lawsuit alleges rent price-fixing by companies using YieldStar software
- Canada investigates rent cartel allegations
- How an algorithm may be helping Canadian landlords coordinate rent hikes
AIID : Incident #894
Historique des modifications
| Version | Date | Modification |
|---|---|---|
| v1 | 8 mars 2026 | Initial publication based on AIID cross-reference scan |