'The Election Was Rigged': Peer-Reviewed Study Confirms Tech Algorithms Systematically Pushed Trump to Victory

A peer-reviewed study published in Nature found that TikTok's recommendation system systematically favoured Republican-leaning political content during the 2024 US election, with researchers at New York University Abu Dhabi saying the effect was strong enough to raise fresh questions about how algorithmic feeds shape political exposure online.

The findings do not prove the election was 'rigged,' but they do show that the platform's feed was not neutral in the months before Americans voted.

The news came after years of concern that social media platforms were doing more than hosting political debate, and instead quietly deciding which side of that debate users saw most often.

In this case, the researchers ran 323 audit experiments using controlled 'sock puppet' accounts across three US states, then tracked more than 280,000 recommendations over 27 weeks leading up to the election.

Rigged Claim Meets Hard Data

The headline accusation may be blunt, but the study itself is narrower and, in some ways, more unsettling. It found that accounts seeded with Republican content received about 11.5 per cent more co-partisan material than accounts seeded with Democratic content, while Democratic-seeded accounts were exposed to about 7.5 per cent more cross-partisan material, largely anti-Democratic posts.

That asymmetry matters because the recommendation system was not simply reinforcing what users already liked. According to the researchers, the imbalance held even after adjusting for engagement metrics, which suggests the algorithm was doing something more than passively reflecting user behaviour.

One of the authors, Talal Rahwan, said the system was not just giving people what they wanted, but was feeding one side more of the other side's attacks.

There was also a sharper edge to the findings when the team looked at candidate-specific content. Trump's official TikTok videos were shown to Democratic-leaning users 27 per cent of the time, while Kamala Harris's videos reached Republican-leaning users only 15.3 per cent of the time.

This is the kind of gap that can look small on paper and still feel pretty mad once you think about the scale involved.

The Narrative And What Social Media Shows

The argument is that TikTok was not an outlier, but part of a wider information system shaped by opaque recommendation engines.

The same piece points to a separate Nature field experiment on X that found users assigned to an algorithmic feed shifted towards more conservative positions, including on policy priorities, Trump-related investigations and the war in Ukraine.

It also highlights a Queensland University of Technology finding that Elon Musk's endorsement of Trump on 13 July 2024 coincided with a structural break in his account metrics, with view counts on his posts jumping 138 per cent and retweets rising 237 per cent.

Those numbers are cited in the source article, though the underlying platform dynamics are still contested and not fully visible to outsiders.

The broader point is hard to ignore. If recommendation systems are steering millions of users towards one side's messages, or away from another side's, then the old idea of a shared political square starts to look a bit flimsy.

TikTok matters especially because, as the source article notes, roughly half of users under 30 use it for politics and news, making it a major route into the election conversation for younger voters.

The Debate Still Isn't Settled

Still, one important distinction needs to stay in view. The Nature study does not say the 2024 election outcome was illegitimate, nor does it prove a coordinated plot to fix the result.

What it does show is that a major platform exposed users to partisan material unevenly, and did so at a politically sensitive moment when many voters were forming views quickly and often through a feed they could not see inside.

That is where the real story sits, not in easy slogans. Algorithms are not election officials, but they can shape what people hear, what they fear and what they come to believe before they reach the ballot box.

If that sounds uncomfortable, it should. The machinery of politics has moved into the feed, and the feed does not care who gets hurt.