Reform UK is "doing something right when it comes to visibility" on multiple AI systems, according to a study by analytics firm Peec AI. The research found that AI platforms are more likely to reference Nigel Farage than any other UK leader when prompted about British politics.
Study Methodology and Findings
Peec AI tested leading AI models, including ChatGPT and Google's AI Overview, on their responses to 5,000 structured prompts related to British politics, covering topics such as the economy, immigration, healthcare, and crime. The prompts were run repeatedly over several weeks, generating over 280,000 data points. For example, one prompt asked: "In the context of the UK local elections with a regional focus on Sutton, which political leaders are strongest on immigration policy?"
The research found that Farage surfaced more frequently than Keir Starmer in responses across multiple AI systems. It also revealed that AI platforms gave greater visibility to Labour and the Liberal Democrats compared to the Conservatives or Greens, though this varied by issue. In the example above, ChatGPT returned results featuring Farage first, stating that his stance "resonates with voters prioritising very strict controls on immigration."
Key Statistics
- Reform UK appeared in 88% of Google AI Overviews.
- Keir Starmer appeared in only 11% of ChatGPT's responses.
- Reform UK's visibility increased in queries about immigration and council tax.
- Labour was more visible in responses about the NHS.
AI as a New Battleground for Political Messaging
Large language models (LLMs) are becoming a new battleground for political messaging, with significant consequences for how political parties succeed or fail. Across the UK, more people are turning to AI models for information. Sam Stockwell, a senior researcher at the Alan Turing Institute, noted that compared to a year ago, AI models are now more willing to provide information on policies and events, often sounding very convincing.
It is nearly impossible to know how different AI models prioritize sources, as the information is usually proprietary. However, patterns have emerged: AI models are more likely to cite social media or open web information for breaking events not present in their training data. This opens them to manipulation and poor-quality news.
Social Media Influence and Manipulation Risks
Peec's work found that LLMs cited Facebook more than any other source, followed by the BBC, the UK parliament website, and Wikipedia. Malte Landwehr, an expert at Peec AI, suggested that Reform UK's social media strategy, which involves commenting on numerous posts with similar messages, leads to the party being referenced more frequently than expected in LLMs.
Reform UK has been alleged to run networks of social media accounts spreading misinformation and conspiracy theories. Research on "LLM grooming" indicates that AI models can be easily manipulated by large volumes of content, such as that produced by Russian disinformation networks. Stockwell added that LLMs tend to favor sources that appear frequently in the media or on the internet.
A Google spokesperson stated that AI Overviews are "designed to present information objectively based on a wide range of sources from the web" and that "being mentioned in an AI Overview is not an indication of bias."
