Simon Popple
School of Media and Communication, University of Leeds. Academic Lead Digital Creativity and Cultures Hub, Senior Lecturer in Photography and Digital Culture. Popple’s research is focused on digital archives, generative AI photography, conflict imagery and digital storytelling.
UK Election 2024
Section 8: Personality politics and popular culture
87. Ed Davey: Towards a Liberal Populism? (Dr Tom Sharkey, Dr Sophie Quirk)
88. Why Nigel Farage’s anti-media election interference claims are so dangerous (Dr Lone Sorensen)
89. Nigel Farage and the political circus (Dr Neil Ewen)
90. Binface, Beany and Beyond: humorous candidates in the 2024 General Election (Prof Scott Wright)
91. What Corbyn support reveals about how Starmer’s Labour won big (Prof Cornel Sandvoss, Dr Benjamin Litherland, Dr Joseph Andrew Smith)
92. “Well that was dignified, wasn’t it?”: floor apportionment and interaction in the televised debates (Dr Sylvia Shaw)
93. TV debates: beyond winners and losers (Prof Stephen Coleman)
94. Is our television debate coverage finally starting to match up to multi-party politics? (Dr Louise Thompson)
95. Tetchiness meets disenchantment: capturing the contrasting political energies of the campaign (Prof Beth Johnson, Prof Katy Parry)
96. “We’re just normal men”: football and the performance of authentic leadership (Dr Ellen Watts)
97. ‘Make the friendship bracelets’: gendered imagery in candidates’ self-presentations on the campaign trail (Dr Caroline Leicht)
98. Weeping in Wetherspoons: generative Al and the right/left image battle on X (Simon Popple)
99. An entertaining election? Popular culture as politics (Prof John Street)
100. Changing key, but keeping time: the music of Election 2024 (Dr Adam Behr)
101. Truth or dare: the political veracity game (Prof John Corner)
Note: Some of the images linked in this text feature offensive racial and gendered representations. The ephemeral and reputational nature of X inevitably means that images will be removed, accounts suspended, and links may become redundant.
This was the first generative AI election in the UK and one marked by the use of tools such as Open AI’s GTP-4 which allows users to create “exceptionally accurate images”. The election saw the first official AI candidate, AI Steve and was foreshadowed by fears of widespread disinformation through deepfakes in the press. CETaS (Centre for Emerging Technology and Security) produced a briefing paper and the Electoral Commission issued advice for voters on the threats of generative AI saying, “We encourage all voters to think critically about what you see and hear in this campaign”.
Whilst these threats failed to materialise, there were limited incidents such as attempts to discredit Wes Streeting over Gaza and Diane Abbott. The major parties eschewed the wider use of AI tools and stuck to traditional approaches to digital photo editing, for example in the Labour Don’t wake up with 5 more years of the Tories campaign. Satirical images remained fixed within established traditions such as photomontage as exemplified by coldwarsteve. Yet there was a discernible nervousness by platforms about the potential of AI images to go unrecognised. An AI-tagged post linking Nigel Farage with Vladimir Putin on Facebook spurred a discussion between two users about how the platform had incorrectly flagged Cold War Steve’s work as AI generated.
One of the low points for the Conservative campaign, and the subject of the first social media-pile-on was Sunak’s early departure from the D-Day commemorations on 6th June. Far right activist Tommy Robinson marked the anniversary with a generated image showing troops storming the beaches under the title “We will remember them”. Unfortunately, the image showed troops storming into the sea with a resulting backlash and a community notice explaining that it was an AI image showing troops retreating at Dunkirk.
Another constant feature of rightist posts was the mobilisation of nationalistic signifiers and widespread use of Lions and Union Jacks to create a series of Narniaesque images by supporters of Reform UK. A common trope was the representation of Nigel Farage surrounded by a pride of union jack bedecked lions. One independent candidate, standing in Wavertree, generated fantasy images to portray the sitting Labour MP as a witch with a frog as a familiar in a series of episodes about their quest for power. Again, this type of image was contested, with a post asking “AI to show me Britain in 2025. Scary”.
Islamophobic tropes were particularly prevalent in the far right’s campaigning and constructed around building fears of invasion, occupation and replacement. The GB News reporter Darren Grimes regularly turned to AI generated images to push phobic images of Muslims and trans people as part of the culture wars. His posts attack migrants and their perceived enablers but generate critical and pointed responses. His use of an image to mimic an official Labour Party ad under the title How has the Labour Party changed? is photo-realistic but again largely countered and its status contested.
Among the counter-narratives developed by the left was a focus on the far-right bogey figure of the so-called ‘Gammon’- the sunburned, overweight beer swilling white male. This archetypal and pejorative image features two weeping men in Wetherspoons and appeared in support to a stream from Anti-Brexit Campaigner Steve Bray who was harassed whilst demonstrating in Jacob Rees Mogg’s constituency. There were many variants on this figure, all drawing on notions of inadequacy, racism and misogyny.
These brief examples illustrate the use of generative AI images outside of the main parties on X and identify some emergent practices, including the use of national and nationalistic signifiers, historical and literary references, user critiques of AI and the counter use of generative images. As post-election stories emerge, such as allegations that Reform UK fielded an AI generated candidate, we need to recognise the longer term impacts of these tools that have accelerated processes to produce ‘political’ images. The rapidity of responses and use of visual posts is noticeable, as are the critical discussions between posters about AI and its implications.
Underlaying this is the implicit need to recognise that data bias is inbuilt into AI foundational models, meaning images have an implicit white western perspective and ignore the global majority. These tools are predisposed to reflect colonising attitudes, and exposing and countering these factors is now an ongoing concern in broader contexts . The introduction of generative AI tools has significantly lowered the bar for rapid image generation and the production of fake, propagandist, and ‘satirical’ images. The potential for misrepresentation is evident and as states mobilize future responses the election has shown the need for awareness and ongoing scrutiny.
