UK Column News — 12th August 2026 artwork

UK Column News — 12th August 2026

UK Column News

August 13, 2026

Mike Robinson, Charles Malet, and Vanessa Beeley present today's UK Column News. Mike Robinson, Charles Malet, and Vanessa Beeley present today's UK Column News.
Speakers: Mike Robinson, Charles Malet, Vanessa Beeley, Brad Miller, Elon Musk

Topics: News

**Mike Robinson** (0:08)
Good afternoon, it's Wednesday the 12th of August, 2026, just after one o'clock. Welcome to UK Column News, your host Mike Robinson. Joining me by video link today are Charles Malet and Vanessa Beeley. Later in the program, we'll have more on the Node Project and its apparent possible drive for Civil War work. I'm going to be covering the Burnham attempts to bribe local communities into accepting electricity infrastructure in the UK. And we're going to be asking what now for Russia in Syria. But we're going to begin today with the continuing rollout of live facial recognition. Because British Transport Police has extended its live facial recognition trial from national rail stations into the London Underground with the first deployment at Victoria Underground Station yesterday. Now this was announced in a police or an arm, sorry, excuse me, a force press release on the 10th of August. The trial is going to rotate between the Underground and network rail sites, and it has been extended to run until November. British Transport Police describes the move as the product of months of work with Transport for London, the Rail Delivery Group, the Department for Transport and the Mayor's Office. And their officer in charge, Chris Casey, said that deployments were going to be intelligence-led and aimed at people wanted by police or the courts, including those breaching bail conditions or court orders. And they said that the technology would be used lawfully, proportionately, and transparently.
TFL's Director of Security, Policing and Enforcement said that live facial recognition had potential for identifying those wanted for high-harm offenses and that the four-month trial is going to be monitored and assessed. But according to their own, that's British Transport Police's own policy material, cameras in these stations are going to be streaming footage to the live facial recognition system, which compares faces against a pre-built watch list.
A possible match is going to generate an alert. An officer then compares the image to the person in front of them before deciding whether to intervene.
They say that images generating an alert are deleted within 24 hours. And they say that biometric data from those who don't trigger an alert is deleted immediately, although the underlying CCTV footage is retained for 31 days. They're going to have signs up and they're going to have QR codes placed around the place so that you can get more information. And they say that alternative routes are available for anyone preferring to avoid the recognition of an alert. But according to their deployment register, across 18 deployments between the 11th of February and the 17th of July this year, at London Bridge, Waterloo, Houston, King's Cross, St. Pancras, Liverpool Street and Victoria, they scanned 537,697 faces. They said that watch lists ranged from 381 to 546 people in each of those cases. And they said that the matching threshold was set at 0.64.
You can decide whether that is of interest or not. But anyway, the point here is, the interesting part is that only one alert was created in those six months at King's Cross in February. And that was a alarm. It was a positive. So there are no confirmed true alerts. There have been no arrests and no recordings of any other kind of outcome. Now, Big Brother Watch, of course, is kind of leading the way in terms of the campaigning on this. They've called the expansion disturbing and dystopian. And they are concerned that the situation, this new infrastructure is treating the traveling public as suspects. Their advocacy manager, Matthew Feeney, said that subjecting law abiding passengers to past biometric surveillance is disproportionate. The language, I think, isn't quite strong enough. And, well, Charles, I mean, I'd be interested in your thoughts on this, because my first thought is I'm not seeing the same kind of pushback on this particular issue that we saw on digital ID, for example.

**Charles Malet** (4:32)
Good afternoon, Mike. Good afternoon all. And I think you're quite right, but there are many issues with it, some of which have been covered. I think one of the reasons that it's not had so much pushback is that people by and large don't really understand it. And one thing to make very clear is that it changes in a fundamental sense, the relationship between the police and the public, in that if a police constable asks you who you are or what you're doing, you are under no obligation to answer that question. And it's only if a suggestion that you're involved in a criminal activity that that can be pushed further.
Facial recognition switches that the other way around, in that police or indeed potentially anybody else is immediately aware of your identity. Now, there are huge implications for that. The other thing to say, which of course has been highlighted by what you've just said, is that if goalposts move with regard to certain legislation, then people may be captured by virtue of who they are. And you look at terrorism legislation where an arrest may be made of a terrorist. So if somebody is described as either a terrorist or a something as yet to be defined, then facial recognition is already a gateway into capturing those people. The other thing to say is that Oxford Street, which I think you might be about to go on to talk about, is also the subject of a redevelopment which is going to have, we believe, trees all over it. So how that's going to work is another question.

42 more minutes of transcript below

Thousands of transcripts fetched by people building searchable podcast archives

Feed this to your agent

Try it now — copy, paste, done:

curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Works with Claude, ChatGPT, Cursor, and any agent that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.

Using your own key:

curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/YOUR_EPISODE_ID