Balancing Innovation and Control: Public Views on Risks and Governance of AI in Political Communication

ABSTRACT With the rise of AI, concerns about its potential for manipulation in political communication have intensified, accompanied by discussions about regulatory measures. Research has explored perceptions of risks and preferences regarding AI use and regulations among the general population, and in fields such as journalism. However, little is known about attitudes toward AI in political communication or how individual factors, such as internal and external political efficacy, may shape these views. Using a representative survey of the German population ( N = 1991), this study aims to assess citizens′ perceptions of risks, acceptance, and regulations for AI use in political communication, and predictors of these attitudes. Results show that AI is perceived as risky and is associated with threats to democracy. AI tasks such as translations are more accepted than its use for creating realistic visual content. Transparency measures are preferred over stricter regulations. Knowledge and curiosity about technology are associated with a more nuanced perspective on regulations, favoring transparency. Fears about AI increase, while external political efficacy decreases preferences for regulatory intervention. Findings underscore that the use of AI in political communication requires education about AI, transparency, and targeted regulatory measures to address concerns and build acceptance for its use.