Two weeks ago, TfL issued London’s first robotaxi passenger licence — and it went to Uber, not Waymo. Since then, the fuller picture has come into focus: how narrow the approval actually is, what Uber and its partner Wayve have already said they want next, and one detail — a hard cap of 15 vehicles — that sits awkwardly next to TfL’s long-standing position that it has no legal power to limit the size of any PHV operator’s fleet.
What was actually granted
On 5 August, TfL granted Private Hire Vehicle licences to a small fleet of Wayve’s autonomous Ford Mustang Mach-Es — electric vehicles fitted with Wayve’s camera-and-radar “AI Driver” system, which notably uses no LiDAR and no pre-built street maps, a different technical approach from most rivals. The licence completes what TfL and Wayve both describe as the “triple-lock” requirement for any private hire trip in London: the operator (Uber, already licensed), the driver, and the vehicle must all separately hold valid TfL licences before a single fare can be collected. A trained, TfL-licensed safety driver remains on board every trip, responsible for the vehicle at all times and able to take the wheel if needed. TfL has been explicit about the category this sits in: the vehicles are “licensed as PHVs, not as automated vehicles” — no new licence type was created. Fully driverless operation, with no safety driver at all, would require a separate approval from DVSA under the Automated Vehicles Act 2024, which has not been sought or granted.
The approval covers up to 15 vehicles, for a one-year trial. Uber and Wayve say the first supervised passenger trips will begin this summer, initially limited to people who registered interest — more than 100,000 Londoners did so in the eight weeks before the announcement — ahead of any wider public launch.
Uber first, ahead of two rivals who got there earlier
The order matters. Waymo has had a roughly 100-vehicle test fleet — Jaguar I-Pace cars fitted with 29 cameras, 6 radars and 5 LiDAR units — operating with safety drivers across London since March, well before Wayve’s tests began, but it does not yet hold a passenger-carrying licence. Waymo has now stated it is aiming for a fully driverless commercial launch in the fourth quarter of 2026, working with Moove as its London fleet operator — the same company confirmed elsewhere as Waymo’s operator in Phoenix, Miami and Las Vegas. Baidu’s Apollo Go, deployed through Lyft’s FreeNow platform, brings by far the largest global track record of the three — more than 22 million cumulative rides worldwide on its RT6 vehicles, which carry 38 sensors including 8 LiDAR units — but its London right-hand-drive testing, which began in Brent on 28 July, remains in early validation, with fare-paying rides not expected until 2027.
Uber, the most recent of the three to start road testing in London, is the first to reach paying passengers. Whatever the regulatory reasoning, one immediate effect of TfL’s approval is that London becomes one of the first cities anywhere where a US and a Chinese AV operator will compete head-to-head for the same market — Baidu via Lyft’s FreeNow, and, in the background, Waymo’s own approach once it clears TfL’s process.
A cap TfL has always said it doesn’t have the power to set
Here is the detail in this approval that deserves more scrutiny than it has had. TfL has told London’s PHV trade for over a decade that it has no legal power to limit how many private hire vehicles a licensed operator can put on the road — that licensing law obliges it to grant a licence to anyone who meets the statutory requirements, with no scope to cap fleet size. That position has been the standard answer given to drivers who have watched Uber’s conventional fleet grow without limit while fares and earnings came under pressure.
Yet in approving Wayve’s vehicles, TfL set a hard limit of 15 vehicles for one year. TfL has not publicly reconciled these two positions. The likely explanation is that this cap flows not from ordinary PHV licensing law but from the government’s AV Trialling Code of Practice, under which these supervised trips also operate — a framework that is explicitly experimental and time-boxed in a way conventional PHV licensing is not. If that is the correct explanation, it would mean these vehicles are being treated as something meaningfully different from an ordinary PHV in practice, even while sitting in the same licence category on paper. TfL has not said this plainly. The fair question, not yet put to TfL on the record: what is the specific legal basis for capping this fleet at 15 vehicles, does that basis differ from the powers TfL says it lacks over the conventional PHV fleet, and if a numeric cap is achievable here, why has equivalent power never been sought or used for the roughly 94,258 conventional PHVs already licensed in London?
What hasn’t moved at all
Driver displacement remains untouched by any of this. TfL has told LDV directly that it holds no projection of licensed PHV driver numbers over the next five years and considers it “too early to make firm predictions” about how many AV operating licences it expects to issue. GMB’s regional organiser, Matthew Wright, responded to the Uber/Wayve licence specifically by urging TfL to exercise “extreme caution” in deploying the technology — language that sits uneasily next to a fleet cap TfL itself hasn’t fully explained the legal basis for, applied to a technology whose promoters have already announced plans to go fully driverless at global scale.
Fifteen vehicles, each with a safety driver, displaces no one today. The real test — for driver livelihoods, and for whether TfL’s “no power to cap” position survives contact with its own actions — comes at the next decision: when DVSA is asked to approve fully driverless operation, when Waymo’s Q4 2026 target arrives, or when the Uber-Wayve-Stellantis ambition for global-scale driverless deployment reaches London specifically. None of those decisions have been made yet. All of them are now closer than they were two weeks ago.
This article updates LDV’s earlier coverage of TfL’s 5 August 2026 licence approval with reporting from Wayve’s own press materials, TechCrunch, Silicon UK, Tech Digest, Newmobility.news and Zag Daily, and incorporates TfL’s response to LDV regarding driver-number projections and AV licensing volumes. LDV has asked TfL to clarify the legal basis for the 15-vehicle cap and will update this article with any response.