How Grok AI thinks the 2028 US election could play out

What happens when you ask an AI model to forecast a future US presidential election? In this breakdown, we look at how Grok, Elon Musk’s AI model, maps out a hypothetical 2028 matchup between Republican JD Vance and Democrat Alexandria Ocasio-Cortez (AOC)—and why it expects a landslide.

The hypothetical 2028 matchup Grok was asked to predict

To build its 2028 election map, Grok was given a specific scenario: JD Vance as the Republican nominee facing Alexandria Ocasio-Cortez as the Democratic nominee. The setup is based on prediction market odds, where Vance and AOC are both seen as plausible future contenders, though not guaranteed nominees.

Grok then projects how each state might vote, assigning them to categories based on expected margins:

• Safe: 15+ point margin

• Likely: 5–15 point margin

• Lean: 1–5 point margin

• Tilt: essentially a toss-up, but with a tiny edge to one side

The result is a full electoral map prediction, with Grok giving Vance 326 electoral votes to AOC’s 212.

Safe Republican states: Grok sees a red foundation plus some

Grok starts by locking in the traditional Republican strongholds as safe for Vance. These include Utah, Wyoming, Idaho, Montana, both Dakotas, most of Nebraska (except its 2nd district), Kansas, Oklahoma, Missouri, Arkansas, Louisiana, Mississippi, Alabama, Tennessee, Kentucky, West Virginia, Indiana, South Carolina, and Alaska.

But the model goes further, adding several states that have become more reliably Republican in recent cycles—or that it believes will be especially favorable to Vance in 2028.

Ohio and Iowa: from swing states to solid red

Ohio and Iowa are two of the clearest examples of long-time battlegrounds that have shifted right. Trump flipped Ohio in 2016 and expanded his margin in 2020 and 2024. With Vance’s home-state advantage and Midwestern profile, Grok projects Ohio as safe red by 2028.

Iowa shows a similar story: Obama won it comfortably in 2012, but by 2024 Trump was winning it by double digits. Grok expects Vance to push that margin past 15 points, also making Iowa safe Republican.

Texas and Florida: firmly in the GOP column

Texas and Florida have both moved from nail-biters to strong Republican territory in recent elections. Trump’s margin in Texas grew from about 5 points in 2020 to nearly 14 in 2024. Florida, once the ultimate swing state, shifted from razor-thin results in 2008–2016 to a 13-point Trump win in 2024.

Grok assumes that a more traditional Republican like Vance, running against a polarizing Democrat like AOC, would lock both Texas and Florida in as safe red by 2028.

After all these safe states are counted, Grok has JD Vance starting at 218 electoral votes, just 52 short of the 270 needed to win.

Safe Democratic states: a smaller but reliable base

On the Democratic side, Grok gives AOC a much smaller set of safe states. These include California, Hawaii, Vermont, Massachusetts, Maryland, Washington, DC, and Maine’s 1st congressional district.

These are deep-blue areas where even a weaker Democratic nominee would be expected to win comfortably. With these locked in, Grok puts AOC at 86 electoral votes from safe states.

Likely Democratic states: blue, but by smaller margins

Next, Grok moves to states that it still expects Democrats to win—but by narrower margins than in the Obama or Biden eras. It places the following in the likely Democratic column for AOC:

• Washington and Oregon – still blue, but not by 15+ points with AOC at the top of the ticket.

• Colorado – a state that has leaned Democratic by around 10 points recently, but which Grok thinks AOC would win by only about 5.

• Illinois – once a 17-point Democratic stronghold, but the margin fell to under 11 points in 2024. Grok expects AOC to struggle to improve on that.

• New York – Biden’s 23-point win in 2020 shrank dramatically in 2024. Grok sees continued Democratic underperformance here with AOC.

• Maine at-large – giving AOC two more electoral votes from the state’s overall tally.

• Connecticut and Rhode Island – reliably blue, but not by the blowout margins of the past.

• Delaware – Biden’s home state, where AOC is projected to win by around 10 points.

After adding these likely blue states, Grok’s map moves AOC up to 179 electoral votes, still well behind Vance even before counting likely Republican states.

Likely Republican states: the Sun Belt and beyond

Grok’s likely Republican column is where the projected GOP advantage really opens up. These are states that have either flipped to Republicans recently or have been consistently close but trending red.

Arizona and Nevada: recent flips that stick

Arizona and Nevada both flipped to Trump in 2024 after voting Democratic in 2020. In Arizona, Trump beat Harris by about 5.5 points; in Nevada, he became the first Republican to win the state since 2004.

Grok expects Vance to build on that momentum, projecting both Arizona and Nevada as likely red in 2028, with margins just over 5 points.

North Carolina and Georgia: still leaning right

North Carolina has been a persistent Republican holdout among swing states, going to Trump in all three of his runs. While margins have been relatively close, Grok expects Vance to win by around 5 points, placing it in the likely GOP column.

Georgia is more complicated. Biden narrowly flipped it in 2020, but Trump took it back in 2024 by about 2.2 points. Grok assumes that Georgia voters will favor a more traditional Republican like Vance and projects a just-over-5-point win, again making it likely red.

Maine’s 2nd district: reliably Republican

Maine’s 2nd congressional district has leaned Republican in recent cycles, and Grok keeps it that way in 2028, rating it as likely GOP.

With all safe and likely states counted, Grok’s map shows JD Vance sitting at 268 electoral votes—just two shy of the presidency—before even considering the closest states.

Lean Democratic states: blue, but increasingly fragile

Grok then looks at states where Democrats still have an edge, but only by a few points. These are rated as lean Democratic in an AOC vs Vance matchup:

• New Mexico – Harris won it by about 6 points in 2024, but Grok expects AOC to underperform, dropping it into lean blue territory.

• Virginia – once a Republican stronghold, now usually Democratic. Biden won it by double digits in 2020, but Harris’s margin shrank to under 6 points. Grok forecasts AOC doing even worse, but still narrowly holding the state.

• New Jersey – perhaps the most dramatic shift. Biden’s 16-point win in 2020 fell to a Harris margin of just under 6 points in 2024. Grok expects AOC to push it into true swing territory, but still slightly blue.

• Nebraska’s 2nd district – a small but important blue-leaning pocket that Grok still gives to AOC by a narrow margin.

With these lean Democratic areas added, AOC reaches 212 electoral votes—still below Vance’s total from safe states alone.

Lean Republican states: the former “blue wall” goes red

The most consequential part of Grok’s forecast is its treatment of the Midwest. Wisconsin, Michigan, and Pennsylvania—once known as the Democratic “blue wall”—have become some of the most contested states in modern elections.

From 1992 through 2012, Democrats won all three in every presidential race. Trump’s 2016 victories there were seen as shocking upsets. Biden narrowly flipped them back in 2020, only for Trump to win them again by larger margins in 2024.

Looking ahead to 2028, Grok expects this rightward trend to continue. It points to:

• Vance’s Midwestern profile – as a former Ohio senator and current vice president, he has strong regional appeal.

• AOC’s lack of regional ties – her political base is in New York, with little direct connection to Midwestern voters.

As a result, Grok places Wisconsin, Michigan, and Pennsylvania all in the lean Republican column for 2028.

Tilt states: Minnesota and New Hampshire on a knife edge

That leaves two of the closest calls on Grok’s map: Minnesota and New Hampshire. These are rated as tilt Republican, meaning they’re essentially toss-ups where Vance has the slightest projected edge.

Minnesota: a historic flip?

Minnesota has not voted for a Republican presidential candidate since 1972. However, margins have been tightening. Obama won it by double digits in 2008, but Clinton carried it by only 1.5 points in 2016. In 2024, Harris beat Trump by about 4.2 points, even with Minnesota’s governor on the ticket as her running mate.

Grok interprets this as a sign that Minnesota is highly competitive and could flip under the right conditions. In an AOC vs Vance matchup, it gives Vance a razor-thin edge, putting Minnesota in the tilt GOP category and awarding him its 10 electoral votes.

New Hampshire: small state, big signal

New Hampshire has a history of close races and independent-minded voters. Biden won it by about 7 points in 2020, but Harris’s margin shrank to just 2 points in 2024. Trump also came within half a point of winning it in 2016.

With a Republican governor and a tendency to favor more traditional Republicans, New Hampshire is another place where Grok thinks Vance could narrowly edge out AOC. It rates the state as tilt Republican, adding four more electoral votes to Vance’s total.

Grok’s final 2028 map: a projected Republican landslide

Once all the categories are filled in, Grok’s 2028 forecast looks like this in a JD Vance vs Alexandria Ocasio-Cortez matchup:

• JD Vance: 326 electoral votes

• Alexandria Ocasio-Cortez: 212 electoral votes

That would make it one of the worst Democratic showings in a presidential election since 1988, when Michael Dukakis lost in a major landslide.

Of course, this is just one AI-driven scenario based on current trends, historical voting patterns, and assumptions about candidate appeal. Real-world campaigns, economic conditions, scandals, third-party candidates, and unforeseen events can all dramatically reshape the map.

If you’re curious how AI models think about the future more broadly, you might also like this look at how AI could evolve by 2028 across coding, business, and everyday tools. And if you want to see another example of Grok reasoning through a complex, controversial question, check out this deep dive into how Grok-4 handled Bible contradictions.

As AI systems get better at analyzing data and simulating scenarios, we’re likely to see more election forecasts like this—useful not as guarantees, but as thought experiments that highlight how shifting coalitions and demographics might shape the next era of US politics.