The debate about whether Uber uses rising insurance costs as cover for growing its margins has been playing out in US courts and academic papers for several months. In Canada, the same structural questions land in a different place - not through regulatory filings or class actions, but through the steady accumulation of tribunal decisions, coverage disputes and broker compliance questions that define how Uber's insurance model interacts with the Canadian market.
The research framing the debate comes from two independent sources. The first is a peer-reviewed Oxford University study, published at the ACM Conference on Fairness, Accountability and Transparency in June 2025, analysing 1.5 million trips from 258 UK Uber drivers between 2016 and 2024. The second is a June 2026 analysis by Len Sherman, adjunct professor at Columbia Business School, drawing on decade-long trip histories from three veteran US drivers. Both found the same pattern: Uber's take rate has passed 50% in many markets, and reported commercial insurance fees on individual trip receipts appear to correlate with fare levels rather than actual risk.
Sherman's most specific finding concerns 100 near-identical trips by a single driver on the same route. Uber's reported insurance and operating expenses varied from $13.75 to $50.00. He ran a regression and found that day of week, time of day and service type were statistically insignificant. What predicted the variance: the rider price and the driver pay. Higher fares were associated with higher reported insurance fees. Lower driver pay was associated with higher reported insurance fees still. On trips where both applied simultaneously, the reported costs were higher.
Consumer Watchdog reported in May 2026 that Uber's self-insurance reserves grew from $6.7 billion in 2023 to $12.5 billion in 2025, while approximately $4.1 billion was simultaneously moved to unrestricted cash on its balance sheet.
The Oxford study found that 93 of 114 drivers tracked across the year before and after dynamic pricing were earning less per hour. Uber's median take rate increased from 25% to 29%, with some trips above 50%.
The Canadian market has its own specific intersection with Uber's insurance model, and it runs through the tribunals.
As Insurance Business Canada has reported, the multi-insurer structure of Canadian rideshare coverage - personal auto with one carrier, fleet accident benefits through another - creates specific failure points. In Gupan v. Definity Insurance Company (2026 ONLAT), a driver injured while carrying Uber passengers filed his accident benefits claim with the wrong insurer because no-one directed him to the fleet policy carrier. By the time he reached Definity, the application window had closed. The claim was barred.
And in the LAT's April 2026 ruling in Dondi v. TD General Insurance Company, the question of what happens when a driver fails to disclose Uber activity to their personal insurer has produced inconsistent outcomes - TD's misrepresentation argument failed in that case not on its merits, but because the claimant could not prove his own case.
These cases reflect a coverage architecture that is fragmented by design. Uber carries fleet commercial coverage for accident benefits; drivers are expected to carry personal auto with a rideshare endorsement; and the interaction between them is, as the tribunals demonstrate, regularly unclear to the drivers themselves. As Insurance Business Canada reported on the gig economy's P&C implications, 22% of Canadian adults - roughly 7.3 million people - now do some form of gig work, and standard policies often exclude carrying passengers or goods for a fee unless a specific endorsement is added. Awareness and take-up vary. Claims disputes follow.
What the Oxford and Columbia research adds is documentation of how the insurance line on Uber's trip receipts behaves at the platform level. The specific concern is that reported commercial insurance fees track fare levels - not risk levels - on individual trips. If that pattern holds in the Canadian market, and there is no reason to expect Canadian trips to be priced differently from UK or US ones given that Uber operates a global algorithmic pricing system, then brokers and underwriters pricing rideshare risk need to factor in that the insurance costs Uber reports on individual transactions may not reflect genuine cost allocation.
Jeff Logan of Mitch Insurance has noted that the rideshare market has "levelled out" in recent years as companies have determined effective practices. Uber's platform architecture has stabilized; the products around it - rideshare endorsements, fleet policies, accident benefits coverage - are better understood than they were. What has not levelled out is the underlying economics of the platform itself, which the Oxford and Columbia research suggest have moved substantially in Uber's favour since the introduction of dynamic pricing.
Uber says its true take, after insurance pass-throughs, is approximately 21% globally. What the research cannot yet confirm - because Uber does not disclose trip-level insurance cost methodology - is whether that characterization accurately describes what is happening in the Canadian market or anywhere else.
