Candidates could use AI to listen to voters rather than broadcast messages
The authors argue that AI in US campaigns is currently used for one-way advertising, similar to old media. They point to examples from Japan and Scotland where AI tools help politicians gather and respond to public input. Such approaches could make campaigning more inclusive and responsive.
Current US campaign AI mirrors old broadcast advertising, prioritizing message saturation over dialogue. In contrast, Japan’s Team Mirai party employs an AI interviewer for extended voter conversations, collecting over 300,000 messages across 16,000 interviews. This input has been cited in legislative hearings and led to platform amendments, contributing to the party winning 12 Diet seats.
Other efforts include CrownShy, a Scottish non-profit developing tools for many-to-many deliberation. These projects draw inspiration from Taiwan’s civic hackers. Such “broad listening” tools collect richer public input than surveys, potentially enabling candidates to respond to constituents at scale rather than simply targeting them.
If adopted widely, this approach could shift campaigning from top-down messaging to genuine constituent feedback, potentially increasing voter engagement and trust. However, it may also raise concerns about manipulation or superficial listening if not implemented transparently. The impact on society depends on whether these tools genuinely inform policy or become another engagement gimmick, possibly reshaping how parties prioritize issues and how responsive governance feels to ordinary citizens.