Londoners may be able to book self-driving taxis next year, and that makes this one of the more useful AI stories of the week. Not because everyone should suddenly trust a car with no human judgement, and not because ordinary people are desperate for a robot to do the school run. It matters because this is what AI looks like when it stops being a chatbot on a screen and starts trying to handle something physical, public and slightly risky.
The immediate trigger is a new burst of reporting around British self-driving company Wayve, which has been showing off autonomous driving in London and is working with Uber on public-road Level 4 trials in the capital. Separately, the UK government’s Automated Vehicles Act implementation programme is working towards the wider framework needed for automated vehicles on roads in Great Britain in 2027, with earlier commercial pilots already being prepared. In plain English: the technology is still being tested, but it is moving closer to normal public use.
What has actually been announced
Wayve and Uber said last year that they planned public-road trials of fully autonomous vehicles in London, working with the UK government and Transport for London on permits and approvals. More recent reporting suggests the government still wants self-driving taxi services to begin appearing from 2026 pilot stages before the fuller legal framework arrives in 2027.
That does not mean driverless cabs will suddenly be everywhere by Christmas. The more realistic picture is a staged rollout: limited trials, closely watched routes, safety rules, and lots of scrutiny if something goes wrong. That is a good thing. Transport is not an area where “move fast and break things” should be anyone’s preferred business model.
Why this is different from ordinary AI headlines
Most AI news is still about writing tools, search summaries, image generators or software for office work. A robotaxi is different because the promise is not convenience in the abstract. It is a real service that has to cope with pedestrians, cyclists, roadworks, rain, aggressive drivers, confusing junctions and the general chaos that London treats as perfectly normal.
That is also why this is such an interesting test of how trustworthy AI really is. In chat apps, a mistake can be annoying or misleading. On the road, a mistake can hurt someone. So if companies want the public to accept autonomous taxis, they will need to prove more than novelty. They will need to show good judgement, predictable behaviour and a clear plan for what happens when reality gets messy.
That broader question of trust is turning up across AI already. We wrote recently about why clearer AI safety labels matter at work and at home, and transport is a perfect example of the same principle. People do not just need a clever system. They need to know what it can do, what it cannot do, and who is responsible when it fails.
What ordinary UK readers should expect
For now, the most sensible expectation is not “cheap robot rides for everyone”. It is something more modest. If pilots go well, people in a limited area may get the option to book an autonomous vehicle through a familiar app. It may sit alongside standard Uber journeys rather than replacing them. There may still be a safety operator in the car during early stages. Prices may not be dramatically lower at first, because new systems are expensive to run and supervise.
Even so, there are a few practical reasons this could matter. In theory, self-driving fleets could help cover quieter hours, fill gaps where drivers are scarce, and make some journeys easier for people who do not drive. If the service is genuinely safe and accessible, it could become another option for airport trips, hospital visits or late-night journeys home. That is a much more grounded way to think about robotaxis than the usual sci-fi framing.
There is also a bigger lesson here about how AI is likely to become useful in daily life. The most realistic future is not AI replacing people everywhere. It is AI being used in narrow, supervised ways where it can remove friction or extend a service. That is why the healthiest mindset is still to treat AI as a helper rather than a magical substitute for human judgement.
If that sounds familiar, it should: it is the same point we made in our recent guide to using AI as a helper, not a substitute. A robotaxi may feel more futuristic than a writing assistant, but the underlying question is the same: where does automation genuinely help, and where do people still need to stay in charge?
The sensible concerns are not going away
There are obvious worries, and none of them are silly. What happens when roadworks force a weird diversion? How does a passenger get help if the vehicle stops unexpectedly? Will these services handle wheelchair users, parents with pushchairs, lost property or sudden changes to the route as well as a human driver can? And what data will be collected from the vehicle, its cameras and the booking app?
Then there is the employment question. Black-cab drivers, minicab drivers and delivery workers are not wrong to pay attention when a new automation push arrives wrapped in language about efficiency. It may take years before robotaxis affect jobs in any meaningful way, but people are right to ask who benefits first and who carries the risk while the system is learning.
The ManyHands view
For ordinary readers, the sensible takeaway is neither panic nor starry-eyed excitement. London robotaxis are worth watching because they are a clear sign that AI is leaving the demo stage and trying to become part of a real everyday service. That is more consequential than another chatbot feature, but it is also much harder to get right.
If the trials are careful, transparent and genuinely useful, they could show how AI fits into public life without demanding that everyone become a tech evangelist. If they are rushed, confusing or badly explained, they will reinforce every existing doubt people have about AI being pushed into places it has not yet earned.
Either way, this is the practical question to keep in mind: not whether robotaxis look futuristic, but whether they make ordinary journeys safer, simpler and more reliable for the people who actually have to use them.
