AI in US Elections: How Campaigns Are Secretly Using ChatGPT
Candidates leverage AI for voter outreach despite transparency concerns and policy bans.
1Widespread AI Adoption
At least 39 U.S. congressional campaigns have purchased OpenAI subscriptions this election cycle, according to campaign finance disclosures analyzed by the Washington Post. The Republican National Committee leads political spenders with nearly $9,700 in OpenAI payments. While most use AI for drafting emails, ads, and scripts, campaigns avoid disclosing this due to voter distrust of AI-generated content. Political consultants confirm actual usage far exceeds official reports, creating a transparency gap in democratic processes.
2Policy Violations in Plain Sight
Multiple candidates admitted using OpenAI tools for prohibited activities like generating targeted ads, despite company policies banning political advertising. Tests revealed inconsistent enforcement - ChatGPT initially created fundraising texts targeting specific demographics (like mothers) in July, but refused similar requests weeks later. Campaigns exploit these inconsistencies, using AI for sensitive tasks like policy development while avoiding detection for restricted uses like social media posts.
3The AI Transparency Paradox
57% of political consultants now use AI daily, up from 34% in 2023, per the American Association of Political Consultants. Yet campaigns face a dilemma: AI provides competitive advantages in voter outreach, but studies show voters distrust AI-generated messages. This drives covert usage, with consultants predicting more AI work will shift to Super PACs as elections near, mirroring current practices with negative advertising.
4Regulatory Patchwork Fails Enforcement
Over 30 states have laws limiting political deepfakes, but enforcement remains weak due to limited resources and constitutional challenges. The Post found OpenAI unevenly enforces its own election rules, allowing circumvention. Former Meta executive Katie Harbath compares AI to negative ads - widely disliked but effective. Left-leaning campaigns now feel pressured to match right-wing AI adoption despite ethical concerns.
5From Policy to Practice
Republican strategist Eric Wilson highlights policy contradictions: campaigns can use ChatGPT to develop education policies but can't create promotional posts about them. This arbitrariness pushes campaigns to test boundaries. OpenAI receives little feedback from political professionals about practical policy impacts, resulting in rules that seem disconnected from real campaign needs and workflows.
6The Super PAC Loophole
Political operatives anticipate increased outsourcing of AI-generated materials to Super PACs, which face fewer disclosure requirements. This mirrors how negative ads are currently handled. Anchor Change CEO Katie Harbath notes campaigns rationalize AI use by pointing to opponents' adoption, creating an arms race dynamic. The tactic allows plausible deniability while maintaining competitive messaging capabilities.
7Voter Trust Erosion
Studies consistently show voters distrust AI-generated political content, yet campaigns prioritize effectiveness over transparency. This conflict mirrors broader challenges in political tech ethics. While election-related deepfakes have gained little traction so far, experts warn the 2024 cycle could see more sophisticated AI manipulation as tools evolve and campaigns refine circumvention techniques.
8Democracy's New Frontier
The AI campaign arms race represents a pivotal test for democratic integrity. With 57% of consultants now daily AI users and enforcement gaps persisting, the 2024 election may become the first truly AI-influenced political cycle. This raises urgent questions about disclosure standards, platform accountability, and whether current regulatory approaches can keep pace with rapidly evolving campaign technologies.
Original article: Slashdot — AI in US Elections: How Campaigns Are Secretly Using ChatGPT