AI in elections: who sets the rules? | DW News

AI is increasingly used in German election campaigns in creative but often non-transparent ways, raising ethical concerns about voter manipulation and misinformation. While new EU legislation will mandate labeling of AI-generated content from 2026, currently there are no clear rules or political consensus in Germany on regulating AI use in elections, leaving transparency and accountability lacking.

The use of AI in election campaigns is becoming increasingly common in Germany, with politicians employing the technology in creative ways such as performing bike loops, dropping rap songs, or even fabricating statements attributed to opponents. However, the main concern is not the use of AI itself but the lack of transparency surrounding its deployment. Voters often remain unaware that AI is influencing the content they see, raising questions about the ethical implications of such practices.

David Fischer, an expert who advises political parties and studies election campaigns, conducted a survey across all German states holding elections this year. His findings revealed that while all parties use AI to some extent, many do so without transparency or proper training for their staff. More than half of the parties did not respond to his inquiries, indicating a general reluctance to openly discuss their AI usage. Fischer warns that using AI tools without clear rules or disclosure sends a message that transparency is unimportant.

Examples of AI use in campaigns include a local Green politician creating a bike loop video to promote bike lanes without disclosing AI involvement, a conservative CDU politician producing a rap song about the city budget with only a subtle caption disclosure, and a far-right AfD politician fabricating statements attributed to a state minister. The latter only added an AI disclaimer after media exposure, highlighting the current lack of accountability and regulation in political AI content.

Experts describe the current political campaign landscape as a “wild west,” with no clear guidelines on what AI-generated content must be labeled or how misleading information should be handled. During the federal election campaign a year ago, a research project called Campaign Tracker identified around 1,000 AI-generated posts, but only 13% were properly labeled. This lack of regulation leaves voters vulnerable to manipulation and misinformation.

Regarding regulation, no political party in Germany is currently advocating for clear, shared rules on AI use in campaigns. However, from August 2026, new EU legislation will require all AI-generated content on social media to be labeled, aiming to increase transparency. Until then, the responsibility for setting rules and ensuring ethical AI use in elections remains unclear, raising concerns about the integrity of democratic processes.