AI-Generated Misinformation Targets Spanish-Speaking Voters Ahead of Election

Voting rights groups warn that AI models are spreading inaccurate election information in Spanish, potentially impacting Latino voters.

Two nonprofit news organizations found that AI-generated election misinformation is more prevalent in Spanish than in English.
Voting rights advocates express concerns that AI-driven falsehoods could widen information gaps for Spanish-speaking voters.
Meta's Llama 3 AI model performed poorly, with nearly two-thirds of its Spanish responses containing inaccuracies.
The inaccuracies are significant in states with large Hispanic populations, such as California, Arizona, Nevada, and Florida.
Advocates urge voters to verify information through multiple credible sources to ensure accuracy.