On Election Day in Israel, the democratic act of casting a ballot was shadowed by something far less visible but just as powerful: a sprawling, coordinated assault of misinformation designed to confuse, provoke, and erode trust. A new study from FakeReporter, an Israeli disinformation watchdog, reveals just how ugly that digital battlefield became. Over the course of a single day, false and manipulated content amassed more than three million views across social media platforms, fueled by artificial intelligence, political opportunism, and even foreign interference. The findings are not simply a catalog of deceptive posts; they are a unsettling portrait of how modern elections are being fought in the public square of feeds and stories. Among the most disturbing examples were AI-generated images shared by Prime Minister Benjamin Netanyahu himself, the cruel misuse of the image of Shiri Bibas—the late hostage whose face had already become a symbol of national trauma—and the discovery of an online network linked to Iran. These incidents did not occur in isolation. Together, they formed an ecosystem of manipulation that targeted voters from multiple directions, aiming not just to sway opinions, but to intimidate people, amplify fears, and cast doubt on the legitimacy of the electoral process itself.
One of the most viral pieces of content centered on a photograph that appeared to show long lines of Arab voters at polling stations on Election Day. The image, which was quickly amplified by coalition lawmakers such as Almog Cohen and media personalities including Yinon Magal, generated over two million views on its own. There was just one problem: it was entirely fabricated. According to FakeReporter’s researchers, the image was generated using an AI image tool. In reality, the photo depicted women ascending the Temple Mount in February 2026, not voters standing in line at the polls. The rapid spread of this fake image served a clear purpose—it fed into a narrative of Arab voter mobilization that had already been stoked during the campaign, and it did so with a visual that looked, at first glance, like undeniable proof. Meanwhile, other falsehoods circulated through less visible channels. An anonymous Telegram account called “Israel News Uncensored,” which boasts more than 123,000 followers, spread an unsubstantiated story claiming that senior members of the Ra’am party had transported a truck carrying eight million shekels in cash to hand out to voters in envelopes. Another X account recycled an older debunked claim about Sheikh Mowafaq Tarif, the spiritual leader of Israel’s Druze community, allegedly instructing Druze voters to support Likud. In fact, the message had been distorted from an older statement where Tarif thanked Netanyahu for aid to Druze communities in Syria. These posts may appear scattered, but together they formed a deliberate pattern: muddy the waters, inflame communal sensitivities, and make it impossible for an ordinary person scrolling through their feed to separate fact from fiction.
Beyond voter manipulation, there was a concentrated effort to delegitimize the election itself. Netanyahu amplified a report from Channel 14 alleging that ballots belonging to right-wing parties had been deliberately damaged. Alongside the claim, he shared a photograph of a crumpled ballot and urged voters to check their own slips. The post drew hundreds of thousands of views, feeding fears that the election was being stolen or rigged. Yet the Central Elections Committee quickly intervened, stating that an extensive examination had found no evidence of systematic ballot tampering. Isolated incidents, they said, were being handled immediately and did not reflect any broader conspiracy. The damage, however, was already being done. The retraction never travels as fast as the accusation, and for many voters, the image of a damaged ballot lingered longer than the official clarification. This tactic—using a single, minor incident to suggest systemic fraud—has become a hallmark of disinformation in democratic elections around the world. It does not need to be true to be effective; it only needs to be seen. And in an age where a single post can reach millions in hours, the line between a localized irregularity and a national crisis becomes dangerously easy to blur. The result is a citizenry that no longer knows what to believe, and an election process that feels fragile even when it is functioning properly.
One of the most alarming findings in the report involves possible foreign interference. Researchers identified a website that was circulated widely in WhatsApp groups supportive of the government. The site portrayed Gadi Eisenkot, a prominent political and military figure, as enjoying widespread Palestinian support and actively working to establish a Palestinian state. The website featured manipulated images of Eisenkot standing alongside Palestinian symbols and appearing at pro-Palestinian demonstrations. On its surface, the site looked like an ordinary political attack page, the kind that has become commonplace in modern campaigns. But a technical analysis revealed something far more sinister: the website was linked to TikTok and Instagram accounts that had been flagged as operating from Iran. This connection suggests that foreign actors are not merely watching Israeli elections from afar—they are actively intervening, using the same tools of division and deception that have been documented in the United States, Europe, and elsewhere. The Eisenkot website is a particularly sophisticated example because it did not rely on crude lies. Instead, it wove together real elements of Eisenkot’s political positions with fabricated images and misleading context, making it difficult for even a careful reader to discern the truth. It is a technique that plays on existing political tensions and personal histories, and it represents a new frontier in covert influence operations.
The FakeReporter study does not exist in a vacuum. It reflects a global trend that news organizations and intelligence agencies across the United States, Europe, and Asia have been tracking with growing concern. In the United States, monitoring groups like Graphika and NewsGuard have documented the increasing use of generative AI to push stolen-election narratives and suppress minority voter turnout—tactics that bear a striking resemblance to what was observed in Israel surrounding Arab voter participation and allegations of ballot tampering. In Europe, organizations such as EU DisinfoLab have investigated foreign influence operations involving cloned websites, a tactic associated with the so-called Doppelganger campaign, which mirrors the fake website created against Eisenkot. Meanwhile, Chinese influence networks known as Spamouflage have used virtual avatars and coordinated TikTok accounts to deepen political polarization in democratic countries. The Israeli experience, in other words, is not an isolated case. It is a microcosm of a much larger phenomenon, where artificial intelligence has democratized the ability to create convincing lies, and where platforms have become unwitting conduits for foreign adversaries and domestic actors alike. The technology is evolving faster than the safeguards designed to catch it. Election officials, journalists, and fact-checkers are engaged in a constant game of catch-up, racing to debunk content that can be regenerated and reshared in seconds.
Nir Rosen, the CEO of FakeReporter, summed up the gravity of the situation with a simple but profound observation. “Election Day illustrated the extent to which false and misleading information spreads across social media, and how quickly such posts can reach millions of views,” he said. What these cases have in common, he added, is the use of unsubstantiated information or manipulated content in ways that can mislead the public and undermine its ability to make decisions based on reliable information. “During sensitive periods such as election day, identifying, verifying and exposing false information is especially important.” His words underscore just how much has changed in the last decade. The disinformation campaigns of the past were relatively crude—text-based rumors and doctored screenshots that traveled through email chains and early social media. Between 2019 and 2022, those tactics evolved into coordinated networks of fake profiles using stolen avatars on X and Facebook, amplifying partisan messages and attacking opponents. Today, artificial intelligence has transformed the landscape entirely. It can generate convincing images of lines at polling stations that never existed, or animate a dead woman’s face to deliver a political message she never spoke. The cost of creating such content is virtually zero, and the realism makes it nearly impossible for the human eye to detect manipulation in real time. The challenge ahead is not just technical, but cultural. We must learn to approach what we see online with a measure of skepticism, and to demand that our leaders and platforms take responsibility for the content they amplify. Because when the truth becomes a casualty of the campaign, democracy itself is at risk.