As deepfakes warp elections overseas, US state legislators race to enact laws, with varying penalties, requiring disclosure of the use of AI in political ads
David W. Chen / New York Times :
Context & Ripple Effects
Election-focused deepfake risk had been visible since early warnings that synthetic audio and video could undermine elections, while the growing use of AI-generated campaign ads had already prompted calls for guardrails. This story shows that concern moving from diagnosis to state-level legislative action.
The state-by-state approach matters because it centers disclosure rather than a single uniform national rule. Later California election-deepfake laws illustrate how this policy track can expand beyond labeling toward platform obligations.
First-order effects
- Political campaigns, consultants, and ad makers face emerging state-specific disclosure requirements when using AI in political advertising.
- State lawmakers are setting different penalties, making compliance exposure dependent on where an ad is distributed and which state rules apply.
Second-order effects
- Campaigns operating across states may need to standardize disclosure practices around the strictest applicable rules rather than tailor every creative to a separate regime.
- The patchwork increases pressure for broader legislative alignment; bipartisan proposals to address AI deepfakes were already emerging in Congress in the wider federal push against harmful synthetic media.
Third-order effects
- If disclosure becomes the default regulatory tool, political AI governance may develop around provenance and transparency duties rather than bans on generation itself.
- Divergent state rules could make platforms increasingly important enforcement points, particularly where laws attach responsibility to handling disputed synthetic political content.
The trend: Election integrity is becoming an early proving ground for public-safety AI rules that impose transparency and accountability duties on AI-mediated communications.