Should Portland amend Affordable Housing Code to add prohibition of anti-competitive rental practices including the sale and use of algorithmic devices (add Code…

This ordinance affirms what has been true for over 100 years: price fixing and coordination are illegal practices that undermine free market competition. This is true whether an individual or entity is colluding with competitors in person in a “smoky back room” or through algorithm-driven price fixing services. Price fixing and coordination facilitated by AI and algorithms are particularly pernicious for both renters and landlords seeking access to a fair, competitive marketplace. They leverage access to mass quantities of sensitive competitor data to “driv[e] every possible opportunity to increase price,” and ensure competitors transform into collaborators, “mov[ing] in unison versus against each other” such that the algorithm “as a middleman, and not the free market, determines the price that a renter will pay.”[1] This means that the landlords willing to collude and undercut small landlords relying on legitimate competitive strategies can profit through market manipulation rather than having to compete to offer a superior rental product. The result is inflated rental prices, market distortions, reduced housing supply, and housing instability.This ordinance will prohibit anticompetitive practices driven by AI and algorithms, restoring the market competition critical to addressing the ongoing housing crisis.Background & Justification:AI or algorithmic price-fixing software, including but not limited to certain software products or modules from RealPage (controlling over 80 percent of the commercial revenue management software market) and Yardi, determine rental prices through data aggregation, machine learning models, and predictive analytics. This is problematic for several reasons:First, the data aggregation process is designed to do just that – aggregate data and align or coordinate prices based on machine learning models and predictive analytics. This is the very definition of price collusion - essentially the virtual iteration of it, fixing and coordinating prices with granular competitor data, in near real-time. This is the algorithmic equivalent of gathering the city’s landlords in a conference room to share sensitive competitor data and agree to set and move prices in unison. This hinders the process of landlords competing independently on their pricing decisions and results in a high probability for coordinated rental price increases.Second, the AI software utilized to determine rental price recommendations is an opaque labyrinth. The machine learning and predictive analytics, such as regression models, neural networks, forecasting, algorithmic clustering, deployed by such companies make it very difficult to understand how they exactly work and how price recommendations are being made. More importantly, the software is deemed proprietary and kept hidden from public understanding and regulatory scrutiny.Lastly, beyond the blatantly anticompetitive practices facilitated by these tools, the ambiguity and lack of transparency with respect to how AI software is functioning to fix rental prices has the potential to quickly devolve into AI models favoring certain neighborhoods, demographics, or tenant profiles, even with the use of aggregated, anonymized data.Lawsuits have been filed at the federal and state levels against these companies, including by the U.S. Department of Justice, Oregon Attorney General, and other jurisdictions, citing violations of antitrust laws and harm to tenants.The ordinance is aligned with a current legislative proposal, Senate Bill 722 (2025), which would prohibit residential landlords from using software modules designed to set rents or occupancy rates and allow affected tenants to collect damages up to a specified amount.The ordinance amends Portland City Code Chapter 30.01.200, explicitly banning the use and sale of such software in the City of Portland. This ordinance does not prohibit landlords from using software that provides landlord support services (e.g., accounting or property management), so long as those tools do not facilitate anticompetitive practices such as price fixing or coordination.This legislation supports Portland’s tenant protection policies and reinforces commitments to housing affordability by prohibiting algorithmic price-fixing in rental housing.[1] United States Department of Justice v. RealPage Inc., Complaint, 24-cv-00710

The reviewed sources do not clearly identify every existing policy, service, contract, or obligation that would remain unchanged.

This proposal concerns housing & homelessness and requires action through Portland’s public legislative process.

Impacted Communities: Algorithmic rental price-fixing software can drive rent increases, and low-income renters, Black and Indigenous communities, seniors, and disabled individuals are disproportionately affected by rent increases because they tend to be more rent-burdened.This ordinance helps stabilize housing costs and reduce displacement, benefiting rent-burdened families and historically marginalized communities. Neighborhood & Livability Impacts: Rising rents and corporate price-fixing have contributed to increased gentrification and displacement, challenges with which Portland has struggled, particularly in East Portland and historically Black neighborhoods.By banning algorithmic rental price-fixing, this ordinance ensures that housing affordability remains a priority for all Portlanders. Public Engagement & Testimony: Tenant advocacy groups, nonprofit housing organizations, and legal experts have been involved in shaping this ordinance.Small landlords and independent property owners have expressed support, noting that algorithmic pricing primarily benefits large corporate landlords at the expense of renters and smaller property owners.Expected testimony will include representatives from tenant rights organizations, housing justice organizations, and impacted renters who have experienced drastic rent hikes due to algorithm-driven rent increases.