A recent AI-generated simulation of the 2028 U.S. presidential election has sparked widespread discussion online, after a YouTube creator used the chatbot Grok—developed by Elon Musk’s artificial intelligence company—to model a hypothetical election outcome years before any official candidates have been selected.
The video asked Grok to run a state-by-state forecast based on assumed contenders from both major parties, producing an electoral map, projected vote totals, and probabilistic outcomes. The simulation ultimately projected a Republican victory, with Vice President JD Vance receiving 312 electoral votes compared to 212 for former Vice President Kamala Harris—well above the 270 needed to win the presidency.
On the Democratic side, the AI model assumed a possible field that included several prominent figures. Early polling referenced in the discussion suggested Kamala Harris leading with roughly 32% support in hypothetical primary scenarios, followed by California Governor Gavin Newsom, former Transportation Secretary Pete Buttigieg, Representative Alexandria Ocasio-Cortez, and Pennsylvania Governor Josh Shapiro. Betting markets cited in the video reportedly showed shifting expectations about whether Harris would run again.
On the Republican side, the simulation assumed Vice President JD Vance as the leading contender for the nomination, ahead of figures such as Donald Trump Jr., Senator Marco Rubio, and Florida Governor Ron DeSantis. The model assigned Vance a strong advantage in early primary probability estimates.
The forecast also divided states into “solid” and “likely” categories based on projected margins. It placed much of the Republican base in states such as Texas, Florida, and Ohio, while assigning strong Democratic margins to states including California, Washington, and Massachusetts. Ohio was highlighted as a particularly notable shift due to its increasing Republican lean in recent elections.
While the simulation generated attention for its detailed electoral map, analysts and commentators emphasize that such forecasts are highly speculative. They depend entirely on assumed candidates, current polling snapshots, and modeling choices rather than concrete election data, since the 2028 race is still years away and politically undefined.
