The Political Ad You Just Read Was Written by AI

In 2024, researchers at Stanford found that more than 30 percent of political tweets supporting either U.S. party showed signs of AI assistance¹. These weren’t trolls. They were staffers using tools like ChatGPT to craft replies, slogans, and policy pitches at scale.

In one case, a congressional campaign in the Midwest ran two social media teams, one human, one AI-assisted. The AI group posted twice as often, reached more voters, and got more engagement². The candidate never disclosed the difference.

AI Is Rewriting Political Messaging

Political teams now use generative AI to:

• Write tailored posts for specific groups

• Reword old content for new audiences

• Simulate voter responses before launch

  
  

A Michigan campaign in 2023 found that AI-crafted messages had 18 percent higher engagement than human-written ones². Another study showed that AI could produce 1,000 political posts in 10 minutes, each adjusted for different demographics³.

This isn’t the future. It’s happening now.

No One Has to Tell You It’s AI

There are no U.S. laws requiring political campaigns to disclose when a message is AI-generated⁴. The Federal Election Commission issued voluntary guidance in 2023, but it has no teeth. Disclosure is optional.

This means a voter may be reading dozens of AI-generated posts each week and not know it. Political speech is becoming faster, cheaper, and harder to trace.

The line between propaganda and automation is fading.

#AI #Politics #Elections #Disinformation #Campaigns #Policy #GenerativeAI #LinkedInArticles

References

1. McKay, D. (2024, March 12). How AI helped write your political ads. Stanford Internet Observatory.

2. Silverman, C. (2023, December 10). AI tools boost campaign engagement in pilot study. MIT Technology Review.

3. Leetaru, K. (2023, September 8). AI-generated political content tested in battleground states. Forbes.

4. Reeve, E. (2023, November 5). FEC proposes voluntary guidelines for AI in campaigns. The Hill.