The Center for Countering Digital Hate reports that AI-generated election deepfakes on X have grown by an average of 130% per month over the past year, driven by easily accessible and jailbreak-prone image generators (Midjourney, DALL-E 3, DreamStudio, Image Creator). Testing of 40 election-themed prompts across these tools showed generators produced deepfakes despite stated safeguards in 41% of attempts, with Midjourney succeeding 65% of the time. Social media platforms have not contained spread effectively: flagged deepfakes continue to accumulate millions of views through reshares, and fact-check labels do not reliably appear on duplicate posts. For financial institutions, this signals near-term regulatory pressure on AI vendors around content moderation liability, potential platform-dependent revenue volatility tied to election cycles, and emerging operational costs for provenance and watermarking tooling.
356 other tracked developments touching the same claims.