Robotaxis will change British car ownership one renewal at a time

Start with the car you might stop paying for

Build the ownership comparison around spending a household could actually avoid. Include insurance, servicing, parking and future depreciation where giving up the vehicle would remove those costs. Then price the replacement journeys, including public transport and occasional hire.

Keep the decision tied to a particular vehicle and a particular moment. Compare replacing a car with using alternatives; account separately for any cost of leaving an existing agreement. Test an older, reliable car with free parking on its own terms, just as you would a second car awaiting an expensive repair.

I would prioritise households considering whether to replace a lightly used second car. They could test booked transport while retaining a vehicle for holidays and journeys the service cannot accommodate. That is the adoption hypothesis I would ask operators to prove.

Avoid building the business case around a supposedly representative annual ownership bill. Ask prospective customers to record the journeys they need, identify the costs they can remove and calculate what remains available for taxis.

Then test the calculation against higher peak fares, extra hire costs and journeys outside the operating area. A proposal to replace a household car should survive those conditions before it earns investment on that basis.

A cheaper mile does not guarantee a cheaper fare

Goldman Sachs’s August 2024 analysis estimated robotaxi vehicle driving costs at US$3.13 per mile, projecting a decline below US$1 by 2030 and to US$0.58 by 2040. Goldman Sachs separately identified much higher costs when corporate overhead and research spending were included.

McKinsey’s January 2025 model estimated that total costs for a US city operation with 1,000 vehicles could fall from US$8.20 per vehicle mile to about US$1.30 by 2035, assuming operations at scale. McKinsey’s proposed savings include better vehicle use, cheaper equipment and lower local operating costs.

I would use the Goldman Sachs and McKinsey forecasts to examine possible cost reductions, while requiring separate evidence of the fares British passengers would pay. Neither Goldman Sachs’s vehicle-cost estimates nor McKinsey’s US operating model establishes when replacing a British household car would become uneconomic.

Require an operating budget that accounts for empty mileage, charging, cleaning, maintenance, depots, customer support and investor returns. Ask how any projected saving would translate into passenger fares after those costs.

Test availability alongside price. Compare the cost of maintaining enough vehicles for peak demand with the waiting times customers would face from a smaller fleet. For an ownership-replacement proposition, measure whether the service completes the difficult journeys as well as the routine ones.

My investment threshold would be evidence that households can afford the service and depend on it when they need it. A forecast of lower vehicle costs cannot satisfy that threshold by itself.

Britain has a service to study and a market still to build

Uber’s stated launch conditions include upfront fares, no extra charge for being matched with a Wayve vehicle and the exclusion of airport journeys at launch. Uber also says expansion will depend on demand, readiness and regulation.

Those conditions should define the scope of any evaluation. Test the journeys Uber says its service offers, and avoid presenting that coverage as proof that a household can dispense with a car.

Baidu announced that Apollo Go began London road testing with Freenow by Lyft on 28 July 2026. Baidu says testing starts in Brent with safety operators onboard and that public rides are planned for 2027, subject to regulatory approval.

Keep road testing, supervised passenger journeys and driverless commercial services distinct in investment decisions. Ask each project to supply evidence appropriate to what it currently offers.

Waymo’s updated 2025 year-end account reports 15 million rides during the year, taking its lifetime total above 20 million. I would accept Waymo’s company-reported totals as evidence of passenger use, while requiring separate evidence of car disposals and full operating economics before drawing conclusions about ownership replacement.

Sustainable Bus reported in December 2024 that the CAVForth autonomous bus service would not continue beyond its planned February 2025 project end. According to Sustainable Bus, Stagecoach attributed the decision to passenger numbers being too low.

I would put the CAVForth experience beside the launch announcements when assessing demand. For every proposed service, require evidence that enough people want the journeys offered at a price and frequency the operator can sustain.

The strongest objection is that ownership buys certainty

The strongest objection to my position is that an affordable booked journey may still be an inadequate substitute for a car outside the home.

Evaluate the practical requirements before claiming a service can replace ownership. Check departure times, passenger and luggage space, child seats, equipment, service boundaries and return bookings. Include the consequences of a failed collection or missed shift in the comparison.

For someone carrying tools between unpredictable destinations, test those journeys directly. Do not assume that a lower average fare answers the ownership question.

YouGov’s survey of 1,000 British adults conducted on 14 October 2025 found that 79% expressed little or no trust in driverless taxis. YouGov also found that, with price and convenience held equal, 85% preferred a human driver.

I read YouGov’s findings as a reason to make trust part of the commercial test. Require evidence of repeat custom, safe journeys and dependable help when something goes wrong. Assess accessibility through boarding, travel and assistance as well as the booking process.

This is why I would focus first on households whose actual journeys fit the service. Test the proposition with a central-city resident using public transport and occasional taxis; assess a family regularly travelling beyond the operating area against its different requirements.

My expectation is that repeated experience could make a renewal decision easier. Operators should demonstrate that experience through the school runs, late shifts and other journeys customers would otherwise use to justify another car.

British technology can earn money beyond British fares

Wayve describes a model in which it licenses its AI Driver to carmakers, while its Uber partnership places fleet ownership and operation with Uber. Wayve’s February 2026 announcement distinguishes planned robotaxi trials from supervised driving software for consumer vehicles.

I see Wayve’s licensing model as a reason for UK technology leaders to examine opportunities beyond operating a local taxi fleet. Apply the same investment discipline: identify the customer, the operational result and the evidence that the customer will pay.

Wayve’s announcement does not disclose a royalty on every autonomous mile. I would therefore assess its stated licensing route without assuming a charging model or treating carmaker investment as proof of future deployment volumes.

For CTC readers, I would prioritise software and services that help operators deliver dependable journeys. Ask fleet buyers what they need to deploy and reverse updates, investigate incidents, control access to remote assistance and schedule depot charging.

Treat those needs as commercial hypotheses to validate with customers. Require a product promising less vehicle downtime to demonstrate that result and connect it to service availability.

Existing fleet businesses should assess charging capacity, maintenance turnaround and the requirements of managing vehicles from different manufacturers. Employers considering staff transport should measure completed journeys, late arrivals and costs through a limited pilot before changing car policies.

For each investment, keep the connection to dependable transport explicit. A supplier should be able to show how its work helps an operator fulfil journeys and sustain the service.

Cheaper taxis could make the city more expensive

I would also make the ownership-replacement case conditional on its effect on the wider transport system. Prioritise services that complement public transport and test whether customers are giving up car journeys or moving out of buses and trains.

McKinsey identifies congestion and empty running as risks. In response to those risks, require occupied and empty mileage alongside passenger numbers when evaluating a scheme.

The Centre for British Progress has proposed charging vehicles while they operate in self-driving mode, discussing charges by distance or time and the treatment of empty journeys. Treat the Centre for British Progress’s proposal as a policy scenario when testing future fares, rather than as an enacted tariff.

I favour judging schemes by whether they improve access while supporting public transport. Give station connections and journeys poorly served by existing routes priority in trials, then measure whether they support the case for relinquishing a lightly used car.

Department for Transport statistics recorded 381,100 taxi and private hire driver licences in England on 1 April 2024. The Department for Transport figure is a dated licence count, subject to qualifications about licence holders; it is not a forecast of jobs lost.

Any investment case built on reducing paid driving should include scrutiny of the consequences for drivers’ earnings. Require proposed fleet and support roles to be assessed by number, location, pay and entry requirements before describing them as replacements.

That scrutiny belongs within the same test of a sustainable service. Assess who can use it, who will pay for it and how the transition would affect the people expected to deliver it.

Watch what happens when a car needs replacing

UK technology leaders should make evidence of household substitution a central test of robotaxi investment.

Look for repeated use through ordinary weeks, bad weather and peak periods. Check whether coverage reaches the destinations that matter to prospective customers. Compare fares after introductory offers have ended and test whether demand survives at prices that cover continuing costs.

Ask participating households to keep a journey diary: where the car goes, who travels and what it carries. Price the alternatives for those actual trips, including journeys beyond the service area. For employers, use a limited travel pilot with measured costs and clear arrangements for failed bookings.

Then follow the renewal decision. Record whether a household replaces its car, keeps it longer or decides it can manage without it. Use that outcome to assess the ownership claim.

I expect self-driving taxis to weaken the case for lightly used private cars where they complement public transport and prove dependable. UK technology leaders should invest against that specific proposition and demand evidence at each step.

The result worth backing is a household reaching a repair or renewal decision with enough experience to choose booked transport confidently. Make that choice the commercial milestone.