Merging AI Incidents Research with Political Misinformation Research: Introducing the Political Deepfakes Incidents Database

Political deepfakes are a documented and growing problem: The paper introduces the Political Deepfakes Incidents Database (PDID), a systematic collection of real-world cases where synthetic media has been used in political contexts, demonstrating that this is not a hypothetical concern but an active phenomenon requiring tracking and study.

The problem extends beyond sophisticated deepfakes: The database includes not only high-quality synthetic videos but also "cheapfakes"—less sophisticated manipulated content—suggesting that policymakers and the public should be concerned about a broader range of manipulated political media, not just technically advanced deepfakes.

Addressing political deepfakes requires cross-disciplinary collaboration: The paper emphasizes connecting AI incidents research with political communication and misinformation research, indicating that solving this problem requires expertise from computer science, political science, policy, and communication fields working together rather than in isolation.

Better data and transparency can support multiple stakeholders: The PDID is designed to benefit policymakers, journalists, fact-checkers, and the public by providing concrete information about deepfake prevalence and trends, suggesting that creating accessible databases of incidents is a practical step toward informed regulation and public awareness.

This tool addresses a critical gap in AI governance: By systematically documenting political deepfakes and their characteristics, the database can support evidence-based policymaking and help develop effective countermeasures, media literacy efforts, and technology governance approaches during a period of rapid generative AI development.