AI Chatbots Could Transform Future Elections, Study Suggests

According to a recent study, researchers have discovered that AI chatbots can be more effective than traditional political advertisements in influencing voter opinions and preferences. The study, published in the journals Nature and Science, involved over two thousand three hundred participants who engaged in conversations with chatbots two months prior to the 2024 US presidential election.

The AI models were trained to advocate for either of the top two candidates, Donald Trump and Kamala Harris, demonstrating a surprising level of persuasiveness, particularly on key issues like the economy and healthcare.

Results indicated that Trump supporters who interacted with a Kamala Harris-favoring AI model shifted 3.9 points towards supporting Harris on a 100-point scale, a shift approximately four times greater than the impact of political ads measured during the 2016 and 2020 elections.

Conversely, the AI model favoring Trump moved Harris supporters 2.3 points in his direction. Additional experiments conducted in Canada and Poland showed even more pronounced effects, with chatbots shifting opposition voters’ attitudes by around 10 points.

The study also referenced an earlier analysis of 24 popular large language models, or LLMs, which revealed that all of them exhibited leftist political views when subjected to political orientation tests.

To explore this phenomenon, researcher Rozado fine-tuned GPT-3.5 to create LeftWingGPT and RightWingGPT, demonstrating that the political leanings of AI models can be influenced by their training data.

Furthermore, a study published in Science analyzed 19 LLMs interacting with nearly 77,000 UK participants on over 700 political issues, finding that the most persuasive models shifted opinions by an impressive 26.1 points.

However, as these models became more persuasive, they increasingly provided misleading or false information, raising concerns over the potential impact on democracy. Wynton Hall, Breitbart News Social Media Director, emphasized the implications of AI chatbots as political persuasion tools, noting the risks associated with their biases and potential for misinformation.

He stated that this development serves as a wake-up call for conservatives ahead of the midterm and presidential elections. The findings from the study highlight the significant role AI chatbots could play in shaping political discourse and influencing voter behavior in future elections.