Though the 2028 U.S. presidential election is years away, an online simulation is already sparking conversation. The YouTube channel “Election Time” recently asked Grok, an AI from Elon Musk’s company xAI, to model a hypothetical race.

The projection examined a potential matchup between Kamala Harris and JD Vance. The AI drew on early primary polling, betting market data, and historical voting patterns to estimate how each party’s nomination might unfold.

In the simulation, Harris leads early Democratic polling with 32 percent support, placing her ahead of Gavin Newsom. On the Republican side, Vance dominates with nearly 50 percent, far outpacing Donald Trump Jr. and other potential contenders.

Based on these numbers, Grok suggested both Harris and Vance would likely secure their party nominations if current trends hold. The forecast remains purely speculative, as the primaries are still distant.

The AI went on to generate a full Electoral College map, labeling states as “solid,” “likely,” or “lean” for each candidate. This method mirrors techniques commonly used by political analysts during election cycles.

Vance was projected to carry reliably Republican states along with battlegrounds such as Arizona, Georgia, and Wisconsin. Harris was forecast to maintain strong support in Democratic strongholds including California, New York, and Massachusetts.

After tallying the projected map, the simulation gave Vance 312 electoral votes to Harris’s 212, suggesting a clear Republican victory in this hypothetical scenario. While entirely speculative, the model demonstrates how AI tools are increasingly being used to explore political possibilities and spark online discussion long before voters head to the polls.