Politics
Generative AI lowers barriers for political deepfakes ahead of 2026 midterm elections
Experts warn that the increased volume, velocity, and variety of synthetic media make it easier for bad actors to manipulate voter emotions and erode trust.
The short version
- Generative AI tools from major tech companies are allowing bad actors to rapidly produce and distribute highly realistic synthetic audio, video, and text.
- Experts warn that these deepfakes are designed to exploit emotional urgency and manipulate voter behavior rather than just creating flawless imitations.
- A key challenge for political campaigns is that they must operate on the same platforms as bad actors while often lacking robust cybersecurity tools.
Key facts
- Generative AI tools from companies like OpenAI, Google, and Meta have lowered the barriers to creating and distributing synthetic media, enabling rapid generation of deepfakes without advanced technical skills.[CNET]
- During the 2024 presidential primary, an AI deepfake robocall mimicking Joe Biden's voice was used to urge thousands of New Hampshire voters not to vote.[CNET]
- A photograph released by Senator Mitch McConnell's office during a July hospitalization sparked widespread social media speculation that it was old or AI-altered, although digital forensics experts found no evidence of alteration.[CNET]
- Michael Kaiser of the nonprofit Defending Digital Campaigns noted that political campaigns face difficulties because they rely on the same public platforms used by malicious actors but lack equivalent cybersecurity defenses.[CNET]