West Midlands Police banned Maccabi Tel Aviv fans from the Villa match, and the country immediately did what it always does: it argued about the morality, the politics, and whether the police are secretly Nazis or secretly Hamas, depending on which end of Twitter you fell out of this morning.
Here is the boring truth. It was probably the most pragmatic decision available.
If you genuinely believe there’s a credible risk of serious violence, and you don’t have the resources or certainty to control it, you don’t “stand firm” for the cameras. You remove the flashpoint. It’s standard military logic. Deny the engagement. Don’t offer the enemy a target. Live to fight another day. Not because a football match is a battlefield, but because the risk maths is the same.
That’s not cowardice. It’s force protection. It’s the difference between a controlled operation and an inquiry with a lot of candles and a photograph on the news.
The scandal isn’t that the police made a hard call. The scandal is that they tried to justify it with material that turned out to be a dog’s breakfast, including an AI-generated false claim that should never have made it into anything with a crest on the letterhead. If your decision is driven by uncertainty and capacity, be honest about that. Don’t pretend you’ve got a watertight evidential case when you haven’t.
And this is where the modern stupidity kicks in: the transactional AI trap. Which, as it happens, I’d only just been banging on about a few days ago, because it’s become the default failure mode of modern institutions.
Someone, somewhere, has treated Copilot like a vending machine for facts. Put question in. Get answer out. Paste into a report. Job done. No checking. No primary source. No adult supervision. Just a nice confident paragraph that looks official enough to survive a meeting.
Except AI doesn’t “know” anything. It predicts plausible text. And when you use it transactionally, it will happily invent a supporting example with the same calm authority it uses to recommend a Malbec you should only drink chilled. The result is what we got here: an official justification padded with a hallucination, like a school essay written at 2am by a teenager who hasn’t read the book.
AI didn’t sabotage the process. The process sabotaged itself by treating AI output as evidence.
Then Chris Philp - the cosplay Farage of the Tory Party - wades in, declaring there would have been violence from “Islamist extremists”. Not “there was a risk”, not “there were concerns”, but effectively “it was going to happen”. With no public evidence. No chain of reasoning. No operational detail. Just certainty, served hot.
And here’s the thing. Philp isn’t in government. He doesn’t have access to live intelligence. So when he declares that violence was inevitable, he isn’t briefing the public from the inside. He’s guessing from the outside, using the scariest label available, and presenting it as certainty.
Which is classic Philp. He’s always struck me as the sort of man who is permanently auditioning for his next job. I’m fully expecting him either to challenge Badenoch for the leadership, or to defect to Reform and start talking about “common sense policing” from a lectern in Clacton.
Fine. Let’s take him at his word. If violence was inevitable, how exactly would he have policed it?
Because once you declare inevitability, you inherit responsibility for the counter-plan. And the options are brutally limited.
Option one: remove the target. That means excluding the away fans. It collapses the whole thing because there is no objective. No target, no spectacle, no confrontation.
Option two: remove the crowd. Play behind closed doors. Ban the home fans as well. Reduce it to a televised training session with stewards and a lot of empty plastic seats. That might reduce the risk inside the ground, but it doesn’t stop trouble elsewhere in the city. And if the target is still in Birmingham after full time, you’ve still got the same problem, just with different timings.
Option three: throw vast resources at it. Mutual aid. Public order units. Sterile routes. Transport hub control. Escorts. Rapid arrest teams. A full-scale operation. Eye-wateringly expensive, massively disruptive, and still with no cast-iron guarantee that you’ve stopped the one determined actor who only needs one gap.
That’s the real world. That’s what policing looks like when you’re dealing with a politically charged threat in a crowd of tens of thousands. It’s not solved by saying “extremists” louder.
So yes, banning the Israeli fans “worked”. Of course it did. It didn’t solve extremism. It just stopped this particular match becoming a stage for it. Removing the target reduces the likelihood of confrontation. It denies the hostile actors their moment. Remove the objective and the operation collapses.
A decision can be operationally correct and administratively indefensible at the same time. That’s what happened here.
And this brings us to the simple question everyone is now dancing around: should the Chief Constable be sacked?
On the facts in the public domain, yes. Not because the force made a hard decision under pressure, but because the leadership allowed that decision to be justified with sloppy, inaccurate material that collapsed under scrutiny. That is a failure of competence and a failure of standards. If you can’t trust the evidential basis for a major public order call, you can’t trust the leadership that signed it off.
There is one complication, and it matters. The Home Secretary can say she has “no confidence” until she runs out of breath, but she cannot just fire him. Under the current rules, that power sits with the local Police and Crime Commissioner. So the political theatre will continue, but the actual decision rests locally.
Sometimes the least bad decision is the only one left. But don’t pad it with AI hallucinations and then act surprised when the paperwork collapses in public.
No comments:
Post a Comment