The streets of Israel were quieter than usual on the eve of the national election, but behind the calm, a different kind of battle was reaching its climax. It was not fought with speeches or rallies, but in the endless scroll of social media feeds, in the shadows of fabricated profiles, and in the silent spread of videos that never really happened. For months, voters have been bombarded with content designed to sway their choices — some of it honest debate, much of it orchestrated deception. Now, with polling stations set to open at dawn, Israeli police were making one final, desperate push to protect the integrity of the vote. Their adversary was not a foreign power, but ordinary citizens, influencers, and political operatives hiding behind screens, using every tool of the digital age to twist public perception. The operation, led by a specialized intelligence unit known as Matzpen, had been running for six months, intensifying with each passing day. And on Election Day itself, from the first vote to the last, these investigators would be watching, tracking, and removing the poisonous content that threatened to blur the line between truth and manipulation.
Matzpen is not a unit that usually fights election battles. Its regular mission, within the Israel Police Intelligence Division, is far more grim: monitoring social media for incitement to violence, for calls to terrorism, for hatred that endangers the state. But as the election approached, its focus shifted dramatically, drawing on its expertise in tracking dangerous rhetoric to a new kind of threat — coordinated influence operations designed to undermine democracy itself. A senior official from the unit described the transition in a rare moment of transparency, acknowledging that the preparation had been underway for half a year. “We have been involved in safeguarding the integrity of Israel’s elections for six months now,” the official said. “We are confronting domestic actors who are trying to undermine the integrity of the elections, and we are working around the clock to monitor the internet and identify influence operations aimed at promoting or discrediting a candidate or political party.” The shift is telling: while the Shin Bet, Israel’s domestic security agency, focused its resources on foreign meddling, the police were forced to confront an uncomfortable truth — the greatest danger to the vote’s authenticity might come from within. Disinformation, in this context, is not a distant geopolitical weapon, but a homegrown tool used by those who know the Israeli electorate intimately, its fears, its hopes, its deepest divides.
The methods being deployed are as sophisticated as they are alarming. Police have identified a growing arsenal of tricks used by domestic actors, each designed to create a false reality. Some create fictitious social media profiles, building networks of fake individuals who comment, like, and share in coordinated bursts to manufacture the appearance of grassroots support or opposition. Others go further, impersonating real candidates, publishing statements in their names, and fabricating news stories that can spread in minutes. The newest weapon in this digital arsenal is artificial intelligence: AI-generated videos that show candidates saying things they never said, or entirely invented scenes that look convincingly real. These are not merely amateur attempts at trickery; they are carefully crafted operations designed to exploit the human brain’s trust in visual evidence. The senior official explained that the unit is also actively disrupting “online impersonators and fictitious personas that publish posts, tweets or TikTok videos containing fake news or impersonating candidates running in the election.” The scale of the problem is immense — investigators admit they can only identify and stop a small fraction of the activity. But their primary goal is not to arrest every offender; it is to stop the damage before it reaches too many eyes. As the official put it, “We are currently focused on monitoring the internet to identify and thwart activity across the various online platforms.”
The concrete actions taken so far reveal the contours of this hidden war. On Sunday, just two days before the election, police announced that they had secured the removal of two entire networks engaged in improper activity on X and Facebook. The details were vague — no public information about who operated these networks or which candidates or parties they were connected to — but the speed and decisiveness of the action suggested a high level of coordination. In a separate incident, the Yashar! party, led by former IDF chief Gadi Eisenkot, filed a complaint after unidentified individuals forged a public message purporting to come from the party. Such forgeries are doubly damaging: they deceive the public and they harm the reputation of the targeted party, even if later revealed as false. When investigators identify suspicious coordinated activity, they do not act alone. The police reach out to the cyber department of the State Attorney’s Office, which then contacts the operators of major social media platforms — X, Instagram, TikTok, Facebook — to request removal of the offending content. This process is careful and legally grounded; investigators must establish that the activity violates rules or laws before any takedown. The cooperation is essential, but it is also a race against time. A misleading post can go viral in minutes, and every hour it remains online could influence thousands of undecided voters.
The limitations of the operation are sobering. Police officials openly acknowledge that they can catch only a small fraction of the networks and individuals engaged in disinformation. The sheer volume of content, the anonymity afforded by the internet, and the rapid evolution of tactics make full enforcement impossible. Moreover, the legal framework is not always clear-cut. Impersonation and fraudulent accounts can constitute criminal offenses, but proving intent and establishing a clear violation is often difficult. The police’s immediate priority, therefore, is not prosecution but disruption — shutting down harmful activity before it causes real damage. This pragmatic approach means that many offenders will never face justice, but it may protect the election from the worst effects of manipulation. The senior official emphasized the urgency: “Our work is becoming increasingly intensive as we get closer to Election Day. Tomorrow will be our busiest day, and until the polls close, we will work to thwart online activity by bots seeking to smear a rival party or promote a candidate or political party.” The statement reflects a deeper truth: the fight against disinformation is not about winning absolute victory, but about minimizing the harm. In this digital age, where trust is fragile and anything can be fabricated, the defenders of democracy can only hope to slow the tide, not stop it.
As the long Tuesday unfolded, police analysts sat in darkened rooms, watching the feeds pulse with new content every second. The operation would continue until the last polling station closed at 10 p.m., but even then, the aftermath would be uncertain. The broader global context remains troubling: democracies around the world are wrestling with the same phenomenon, where foreign and domestic actors exploit social media to sow confusion and division. Israel’s election is a microcosm of this larger struggle. For the ordinary voter, the lesson is simple but vital: election content, especially that which provokes strong emotions or confirms preexisting beliefs, must be approached with skepticism. If a video looks too perfect, if a quote seems out of character, if a post appears out of nowhere, pause and verify. The police can do their part, but the ultimate safeguard is an informed and cautious public. In the final hours, as the sun set and the votes began to be counted, the digital battleground grew quiet. The bots were silenced, the fake accounts taken down, the viral content removed. But the underlying threat remains, waiting for the next election cycle. For now, the officials breathed a sigh of relief that the vote had been conducted without a major crisis. Yet the question lingers: in a world where technology outpaces regulation, can any election truly be free from manipulation? The answer, perhaps, lies not in the algorithms but in the hands of the voters themselves, who must learn to navigate the noise and find the truth among the lies.