A new AI-driven simulation of the 2028 U.S. presidential election has been gaining traction online, largely due to a collaboration between the YouTube channel Election Time and Grok AI, the platform developed by Elon Musk’s company xAI. The video presents a hypothetical Electoral College forecast based on a matchup between Kamala Harris and JD Vance, walking viewers through simulated primaries, polling trends, betting markets, and demographic modeling. Importantly, the creators emphasize that this is not a prediction but a scenario-building exercise—essentially a “what if” exploration using data-driven assumptions. The AI model incorporates historical voting patterns, recent electoral shifts, and population changes to construct a plausible map of how the race could unfold. This kind of simulation appeals to audiences because it blends political analysis with emerging AI capabilities, offering a structured way to think about future elections long before they happen.
On the Democratic side, the simulation places Harris as the early frontrunner, with a projected 32% support in primary polling. She leads ahead of Gavin Newsom, who is modeled at 23.8%, while Pete Buttigieg trails further behind. Other figures, including Alexandria Ocasio-Cortez and Josh Shapiro, appear in the simulation but with more limited support, reflecting a fragmented early field. The AI attributes Harris’s advantage to factors like national recognition, fundraising ability, and the historical tendency for well-known figures to dominate early primary stages. Interestingly, the model also suggests a rebound effect following her 2024 loss to Donald Trump, indicating that prior defeat does not necessarily eliminate future viability. Instead, it can sometimes consolidate party support, especially if no clear alternative emerges.
For Republicans, the simulation shows a much more consolidated field, with Vance holding a commanding lead at 49.2% in early polling. He significantly outpaces Donald Trump Jr., who is modeled at 20.2%, along with Marco Rubio and Ron DeSantis, who trail further behind. The model highlights the advantage of incumbency—real or perceived—as a major factor in Vance’s dominance, along with continued Republican strength in key regions following the 2024 cycle. It assigns Vance a 46% probability of becoming the GOP nominee, suggesting a relatively clear path compared to the more competitive Democratic field. Still, as with any simulation, these numbers are sensitive to assumptions; shifts in public opinion, unexpected events, or new candidates could quickly reshape the landscape.
When mapping out the Electoral College, the AI begins by assigning “solid” states—those with large, consistent partisan margins. In this scenario, Vance secures a broad swath of reliably Republican states across the South, Midwest, and Mountain West, including places like Texas, Tennessee, and Utah. Notably, Ohio is classified as solid Republican, reflecting its continued shift to the right in recent elections. Harris, meanwhile, holds onto traditional Democratic strongholds such as California, New York, and Massachusetts, along with much of the West Coast and parts of the Northeast. After allocating these states, the simulation gives Vance an early lead of 139 electoral votes to Harris’s 108, underscoring the structural advantage Republicans currently hold in the Electoral College map, at least within this model’s assumptions.
The gap widens further when “likely” states are added—those leaning toward one party but still somewhat competitive. Vance is projected to win states like Florida, North Carolina, and Arizona, reflecting continued Republican momentum in the Sun Belt. Harris picks up states such as Illinois, Virginia, and Colorado, though the model notes that some of these show reduced margins compared to previous cycles. At this stage, Vance reaches 246 electoral votes, just shy of the 270 needed to win, while Harris sits at 212. This highlights a key dynamic: Democrats may need to win a larger share of true battleground states to overcome the GOP’s broader geographic advantage.
Finally, the simulation turns to “lean” and “tilt” states—the most competitive parts of the map. Here, the model projects narrow Republican advantages in key battlegrounds like Pennsylvania, Michigan, and Wisconsin, echoing patterns seen in recent elections. States such as Georgia and Nevada also lean Republican in this scenario, while traditionally Democratic-leaning states like Minnesota and New Hampshire are classified as razor-thin contests. In the final projection, Vance reaches 326 electoral votes, securing a decisive victory, while Harris’s support remains concentrated in coastal and northeastern regions. The result illustrates how small shifts in a handful of competitive states can have an outsized impact on the overall outcome. At the same time, it reinforces the central point: this is not a forecast set in stone, but a structured thought experiment—one that highlights trends, assumptions, and the evolving dynamics of American electoral politics.
