The 2028 U.S. presidential election is still years away, but a new online simulation is already generating debate. The YouTube channel “Election Time” asked Grok, an AI developed by Elon Musk’s company xAI, to model a hypothetical race.

The projection focused on a potential matchup between Kamala Harris and JD Vance. Using early primary polling, betting market data, and historical voting patterns, the AI estimated how both party nominations could unfold.

According to the simulation, Harris leads early Democratic polling with 32 percent support, ahead of Gavin Newsom. On the Republican side, Vance was shown dominating with nearly 50 percent, well ahead of Donald Trump Jr. and other potential contenders.

Based on these numbers, Grok suggested both Harris and Vance would likely secure their respective party nominations if current trends held. The forecast remains speculative, as the primaries are still far in the future.

The AI then generated a full Electoral College map, labeling states as “solid,” “likely,” or “lean” for each candidate. This approach mirrors methods 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 illustrates how artificial intelligence tools are increasingly being used to explore political possibilities and spark online discussion long before voters head to the polls.