WASHINGTON — The Supreme Court's decision to gut Section 2 of the Voting Rights Act set off a redistricting race across the country, with Republican-controlled and Democratic-controlled states alike moving to redraw congressional maps in ways that benefit their party. The legal fights that follow those maps into court increasingly turn on a single piece of evidence: computer algorithms that can generate and analyze millions of possible district configurations for any given state.
Those algorithms work much like generative AI models such as ChatGPT or Claude. They are trained on millions of randomly generated maps through a complex sampling process, which lets them produce a statistical picture of where any particular legislature-drawn map falls relative to the full universe of plausible alternatives. When a judge asks whether a map was drawn to favor one party, the algorithm can answer with a precise probability.
A Utah case shows what that looks like in practice. In November, a state judge threw out a congressional map drawn by the Republican-led Utah legislature for violating a 2018 voter-approved proposition intended to end partisan gerrymandering. The judge's ruling cited computer-simulated analysis finding that the rejected map was "more Republican than over 99 percent of expected maps drawn without political considerations." The legislature's map did not merely lean Republican — it sat at the extreme tail of every statistically plausible alternative.
The Utah ruling is not an isolated event. Legal and political science experts say algorithmic analysis is now a standard weapon for both sides in high-stakes redistricting litigation, used either to attack a legislature's map or to defend one. As the Supreme Court's VRA decision pushes more states to redraw districts, the number of such cases heading to court is set to rise sharply, in some cases within months.
The stakes extend directly to control of Congress. Redistricting fights in large states can flip multiple House seats with a single court order. Because courts often operate under tight timelines when election cycles are approaching, algorithmic analysis gives judges faster and more statistically grounded evidence than traditional expert testimony.
Supreme Court Justice Elena Kagan made the underlying threat explicit in her dissent in the 2019 Rucho v. Common Cause case, when the Court held that federal courts have no role in policing partisan gerrymandering. "Old-time efforts, based on little more than guesses, sometimes led to so-called dummymanders — gerrymanders that went spectacularly wrong," Kagan wrote. "Not likely in today's world. Mapmakers now have access to more granular data about party preference and voting behavior than ever before."
Kagan's concern has only grown more relevant since 2019. The data available to mapmakers has become more precise, and the AI tools available to both mapmakers and their opponents in court have grown more powerful on the same timeline. A gerrymandering effort that was difficult to detect a decade ago is now measurable to within a fraction of a percentage point against the statistical distribution of all possible alternatives.
The mechanism matters for understanding why these tools are persuasive to judges. A legislature can always argue that its map reflects neutral factors — keeping counties whole, preserving communities of interest, drawing compact shapes. The algorithmic analysis sidesteps those arguments by asking a purely statistical question: across millions of maps that respect all those same neutral criteria, what fraction are as favorable to one party as the map under review? If the answer is fewer than one percent, the neutrality claim becomes difficult to sustain.
State courts have become the primary arena for these fights since Rucho closed the federal courthouse door on pure partisan gerrymandering claims. That makes state constitutional provisions and state-level anti-gerrymandering laws — like Utah's 2018 proposition — the main legal hooks, and algorithmic evidence the main way to prove a violation under those standards.
The asymmetry between mapmakers and challengers that Kagan identified in 2019 now cuts in both directions. Legislatures drawing maps have access to fine-grained voter data and increasingly use their own algorithmic tools to construct maps that are partisan without being obviously so. Challengers, who once had to rely on visual inspection or basic demographic comparisons, now deploy the same class of algorithms to expose what the mapmakers did. The result is an arms race inside the courtroom.
Tyler Simko, an assistant professor whose work on map-simulation methods has fed directly into redistricting litigation, is among the academic researchers driving that shift. The academic literature on algorithmic redistricting analysis has moved from journals into courtrooms faster than most legal scholars anticipated when the Rucho decision came down.
