How AI and Russian Trolls Can Influence Voting Behaviour
amsterdam, zaterdag, 20 september 2025.
Professor Claes de Vreese warns that political parties and Russian trolls could theoretically use AI chatbots to influence voting behaviour. These bots can gather and disseminate information that affects public opinion. Researchers at Brown University have shown that it is possible to steer the political bias of AI chatbots, but the political and ethical risks are significant. The GreenLeft-PvdA is optimising their website for AI tools to spread accurate information, while other parties approach the possibilities of AI with caution.
How AI and Russian Trolls Can Influence Voting Behaviour
Professor Claes de Vreese warns that political parties and Russian trolls could theoretically use AI chatbots to influence voting behaviour. These bots can gather and disseminate information that affects public opinion. Researchers at Brown University have shown that it is possible to steer the political bias of AI chatbots, but the political and ethical risks are significant. The GreenLeft-PvdA is optimising their website for AI tools to spread accurate information, while other parties approach the possibilities of AI with caution [1].
The Risks of AI in the Political Arena
The possibility of influencing AI chatbots to spread political views is both fascinating and alarming. Chatbots can collect information from the internet and use it to influence public opinion. Political parties could theoretically hire Russian troll armies to manipulate AI chatbots, but the political risk is high. Public forums such as Reddit, LinkedIn, and X are vulnerable to manipulation by AI chatbots [1].
Practices of Political Parties
The GreenLeft-PvdA has decided to optimise their website for AI tools so that they can provide accurate information to voters. This is an attempt to ensure the reliability of the information disseminated by AI chatbots. However, other parties are more cautious and are investigating the potential risks and ethical issues associated with this approach [1].
Implications for Media Literacy and Democracy
The rise of AI in the political arena has a significant impact on media literacy and democracy. Voters must be critical when evaluating information disseminated by AI chatbots. It is important to consult multiple sources and verify the authenticity of the information. Researchers note that one in three responses from AI chatbots can be incorrect [2].
Practical Tips for Voters
To recognise fake news, voters can apply the following practical tips: 1) Consult multiple reliable sources, 2) Check the date and source of the news, 3) Look out for grammatical errors and unnatural language use, 4) Seek confirmation of the news in other media, 5) Be critical when reading sensational headlines. By following these steps, voters can be better protected against disinformation [2].