AI in US Politics

Basic Concepts

Q: What are deepfakes?
A: AI-generated videos that falsely depict people saying or doing things they never did.

Q: How are deepfakes spread?
A: Through social media, online forums, and digital ads.

Q: How many deepfakes were detected in 2019?
A: 15,000 (an 84% increase from 2018).

Q: What percentage of Americans believe deepfakes will confuse voters?
A: 63%

Q: What percentage of deepfake targets are politicians?
A: 39.2%

Q: How many deepfake videos of Ron DeSantis were found in 2024?
A: 48 videos with 5.8 million views across 12 platforms.

Proposed Solutions

Q: What are the three main solutions to AI threats in elections?

1. Election-based legislation

2. Mandatory AI detection for campaign ads

3. Voter education on AI and deepfakes

Solution 1: Election-Based Legislation

Q: How can legislation help?
A: Regulates AI use in political ads to prevent misinformation.

Q: Why might legislation be ineffective?
A: Many voters ignore disclosure labels.

Solution 2: Mandatory AI Detection

Q: What is AI detection?
A: A system that scans political ads for deepfakes or misleading content before they reach the public.

Q: What is an example of successful AI detection?
A: MIT’s AI system, which detected disinformation with 96% accuracy and was used in the 2017 French elections.

Q: Why is this the most practical solution?
A: It prevents misinformation before it spreads and is easier to enforce than voter education.

Solution 3: Voter Education on AI

Q: What is the goal of voter education?
A: To help citizens recognize deepfakes and AI-generated disinformation.

Q: What is the biggest challenge?
A: Public resistance due to political bias concerns.

Q: How can education be effective?
A: It must be neutral and transparent to gain public trust.

Source Evaluation

Q: What were the three sources analyzed?

1. Inc.com article on transparency – Shows trust is key but is business-focused.

2. MIT study on AI detection – High accuracy but based on older data.

3. NY Times article on misinformation – Supports voter education but notes its limitations.

Q: Which source is the most reliable?
A: MIT study on AI detection (96% accuracy in spotting misinformation).

Final Conclusion

Q: What is the best solution?
A: Mandatory AI detection for campaign ads.

• It stops misinformation before it spreads.

• It is more enforceable than legislation or education.