Reform UK politicians keep trying to pull the wool over our eyes using artificial intelligence (AI).
Despite the tech improving considerably over the past few years, it’s often still possible to identify AI images at a glance.
The reason for this is that the people operating the technology don’t have an eye for distinguishing between bad images and good ones. Arguably, we should be glad we can still tell the difference otherwise Reform would have had us fooled.
In Plymouth, Reform Cllr Ben Rowe posted an image showing six smiling colleagues who don’t exist. pic.twitter.com/ZENBcJQ2MI
— HOPE not hate (@hopenothate) August 31, 2026
Reform thinks we can’t tell these are AI-generated…
AI images remain uncanny in ways that can make them unsettling to look at. In the image above, for example, everyone seems to have the exact same expression.
The three men don’t just have the same expression either; they all look like variants of the same guy. While you could assume the three men were father and sons, AI does have a tendency to clone people like this. This is very apparent in the following image from Reform Essex.
Reform UK Essex has launched OPERATION VANGUARD.
Putting security personnel on the streets of Colchester, Chelmsford and Braintree to help protect residents and provide a visible presence.
We were elected to put residents first, and that is exactly what we intend to do. pic.twitter.com/MwOUpuFMFT
— Cllr Sam Journet (@ReformUKSam) August 28, 2026
The two men appear to be identical twins, and yet one of them is half a foot shorter. That isn’t how identical twins work.
In the above example, AI is far from the biggest complaint either. As Narinder Kaur said:
Unaccountable “security patrols” with no defined stop criteria or policing powers and no transparency?! Sounds like a recipe for racial profiling. Reform UK Essex needs to explain who decides who gets stopped, and why.
People have also pointed out that Reform UK has a terrible track record when it comes to vetting. As such, they’re perhaps the last people who you’d want to be deploying private security professionals on the streets.
Councillor forgets to remove prompt
In another post highlighted by Hope not Hate, a Reform councillor was caught pasting in an AI response. What made it obvious is she wasn’t supposed to use the first paragraph.
Here’s a Facebook post made by Reform Cllr Sophie Preston-Hall in Essex, in which she forgot to remove the AI prompt… pic.twitter.com/sLVo9uvu8c
— HOPE not hate (@hopenothate) August 31, 2026
Some people might not care if their local politician is using ChatGPT to improve their written text. They will definitely care about the trend of using AI to fake work being done, though.
And the biggest howler: Reform’s geniuses in Kent were apparently able to repaint a fence, trim the verge, plant grass seed (and get it to grow instantly), redo the pavement and repave the road all so quickly that the car in the top-right corner didn’t move an inch! Amazing! pic.twitter.com/Oa5g66VAc8
— HOPE not hate (@hopenothate) August 31, 2026
This sort of stuff gets to the heart of why AI is such a big problem when it comes to politics; namely that it further erodes the trust between citizens and elected officials.
As far as we can tell, Reform UK seems to be leaning on this technology more heavily than other parties too. Notably, a Reform councillor in Swindon resigned on 27 August, citing AI as a primary problem for him.
BREAKING: Reform Swindon councillor quits claiming party uses AI and tells members how to votehttps://t.co/cMFoJvGkHJ
— Reform UK Exposed
(@reformexposed) August 27, 2026
Councillor Lawrence Davies said:
I am no longer confident in the way the party is governed.
I have been a long-standing Reform voter and campaigner, but I cannot with clear conscience continue to fight for Reform UK in Swindon when members are told how to vote and what to say by their officers and whip, and every question and every speech is rewritten with AI.
The ChatGPT Party
The most notorious example of Reform relying on AI was of course, former candidate Matt Goodwin’s ‘MattGPT‘ fiasco.
Analysis of Goodwin’s new book, Suicide of a Nation, found that it invents numerous quotes by historical figures, uses false… pic.twitter.com/ywNgV2AEsw
— Europa.com (@europa) March 27, 2026
Reform politicians may be saving themselves a bit of time by using this technology, but they’re not winning themselves any trust.
Featured image via Yui Mok/ PA
By Willem Moore
From Canary via This RSS Feed.