Uber will trial robotaxis - autonomous cars with no human safety driver at the wheel - in London next spring.

The ride-hailing app will work with the UK^▸ (United Kingdom) artificial intelligence (AI) firm Wayve, which has been testing out the technology on the city's streets with human oversight, in line with current legislation.

The announcement comes after the UK government changed its rules about the driverless cars once again.

It was originally aiming for the tech to come to British roads in 2026, then the date was changed to the second half of 2027. But it now says it is introducing an accelerated framework for small autonomous "bus and taxi like" commercial services to get them underway earlier.

It is not yet clear whether the vehicles in Uber's trial will be available for customers to use - the firm says it is still working out the details. It has previously said it intends to add them as a regular option via its UK app as soon as legislation allows.

The Department for Transport says the industry could create 38,000 jobs and add £42bn to the UK economy by 2035.

But speaking to the BBC last month, GMB national secretary Andy Prendergast said the "significant social implications" driverless cars and taxis could have - including on unemployment - should also be fully considered.

Uber launched a robotaxi service in Austin, Texas in March and said its driverless vehicles could work for 20 hours per day, seven days per week. Customers there can choose whether to take a robotaxi if there is one available, with no difference in fare. Tesla is planning to launch a rival service in the same city in June.

Fully driverless cars have done millions of miles on public roads in other countries too, including China, UAE and Singapore, but whether they are more or less safe than human-driven ones is still being investigated.

Numerous studies suggest automated vehicles are less accident-prone than human drivers, based on US data. But there have been a number of incidents involving robotaxis in the countries where they operate, ranging from road accidents to passengers being locked in. And one service in San Francisco was cancelled after a series of malfunctions.

In May I took a ride in a car fitted out with Wayve's autonomous kit across central London. We had a human safety driver at the wheel but he did not have to use the controls once during our 30 minute journey. The car handled every potential hazard which appeared in the busy streets including congestion, temporary traffic lights, cyclists and, at one point, a pedestrian using crutches in the middle of the road.

The Ford Mach-e was fitted with sensors and a radar, and an AI-powered system controlled the vehicle's responses in real time. If anything it was a lot more cautious than a human driver, which made for a reassuringly uneventful trip.

Not sure I would use one. Who is the responsible driver/person if the car malfunctions/crashes (perhaps owing to the fault of the other side too)? Who handles any claims etc?