An artificial intelligence model generated a hypothetical prediction for the 2028 U.S. presidential election, projecting Vice President JD Vance defeating New York Representative Alexandria Ocasio-Cortez in an electoral college matchup.
The prediction was created using Grok, an AI chatbot developed by xAI, following a request from the YouTube channel Election Time. The simulation analyzed current polling data, betting market information, and historical election patterns to model the potential race.
What the AI Projected
Grok projected Vance winning 326 electoral votes to Ocasio-Cortez’s 212, surpassing the 270 votes needed for the presidency. The model showed Vance securing traditional Republican strongholds across the Mountain West, Great Plains, and South while gaining ground in key battleground states.
According to the analysis, Vance would win Ohio, Arizona, Nevada, North Carolina, and Georgia. The simulation also placed Pennsylvania, Michigan, and Wisconsin in lean Republican territory, representing a significant shift from their recent Democratic voting patterns.
Minnesota and New Hampshire flipped red in the projection, delivering what the analysis described as decisive margins in those states.
How the Prediction Was Built
Grok based its analysis on multiple data sources including current polling, betting markets like Kalshi, and patterns from previous elections. The AI evaluated candidate positioning and voter preferences across demographic groups and geographic regions.
Vance currently holds approximately 32% of Republican nomination odds in actual betting markets, according to the analysis. The projection considered his Midwestern background and current political standing in its modeling.
This story has been updated. CNN’s Digital team contributed to this report.
