In the sweltering early days of August 2026, French internet users scrolling through social media came across something deeply troubling: a video of Edwy Plenel, a well-known investigative journalist, seemingly confirming that his co-host Léa Salamé had bribed major news outlets for positive coverage of her partner, presidential candidate Raphaël Glucksmann. It was the kind of story that feels real because it feeds directly into the public’s worst fears about backroom deals, corrupt media, and powerful people protecting their own. For a few hours, it spread through messaging apps and social platforms, leaving a trail of outrage and suspicion. But it was not real. The video had been generated by artificial intelligence, and France’s national security services later traced it to Storm-1516, a Russian network run by the GRU. This was not an isolated dirty trick. It was a preview of a much larger strategy designed for a French presidential election that had not yet begun in earnest. What makes this strategy so dangerous is not that it is sophisticated. It is actually cheap, repetitive, and deliberately messy. Russian operations are built to be redundant and impossible to fully neutralize. They do not need any single voter to believe a single lie. They need to create enough noise, enough contradiction, and enough suspicion that voters become exhausted and cynical. Even when a fake is exposed, the residue of confusion remains. The fact that officials spent energy denying it makes some people wonder: why would they deny it if there were nothing to it? The goal is not persuasion; it is pollution. Once the information environment is polluted, no amount of cleaning restores the original trust.
To understand why these attacks happen, you have to look at the bigger picture. Russia has spent years cultivating relationships with the French far right and working to weaken public support for Ukraine. France has been one of the most important European voices in favor of continued military and financial aid to Kyiv, and that makes it a target. In the summer of 2026, this strategy shifted into higher gear. Between July and August, three presidential candidates were targeted by Russian-linked networks. Before the Glucksmann-Salamé deepfake, Édouard Philippe and Gabriel Attal found themselves at the center of fabricated health rumors. Philippe, it was claimed, had frontotemporal dementia. Attal was accused of having Parkinson’s disease. These are cruel and frightening allegations, the kind that can rattle not just a campaign but an actual human being and their family. And they were not random. Each of those men has been outspoken in support of continued European military assistance to Ukraine. Philippe has called for stronger backing, Attal served as prime minister under a president firmly committed to Kyiv, and Glucksmann built his entire political career around European federalism and support for Ukraine. Not one of the major candidates sympathetic to Moscow, most notably Marine Le Pen, was targeted in these operations, even though she was recently convicted over the misuse of European Parliament funds and even though legal questions continue to hang over her candidacy. Le Pen is currently leading in first-round polling. The message writes itself: support Ukraine, and you become a target. VIGINUM, the French agency created to counter foreign digital manipulation, has described the strategic objective of these networks in exactly these terms: to erode Western support for Ukraine by making the people who advocate for it seem untrustworthy, sick, corrupt, or simply too risky to elect.
To understand how this works, it helps to lift the hood on the operation. Over the past two years, French authorities have mapped the Storm-1516 playbook in remarkable detail. In a May 2025 report, the agency documented 77 campaigns by this single network between its emergence in August 2023 and March 2025. The stories change, but the method is almost monotonously consistent. First, disposable social media accounts—often bought in bulk from suppliers—appear out of nowhere to post fabricated evidence. The posts are framed as whistleblowing, as if a brave insider has finally decided to expose the truth. Then paid third-party websites and affiliated amplifiers pick up the content and republish it, making it look like independent journalism. Next, real pro-Russian sympathizers, who may have no connection to the original conspiracy, share it further because it aligns with their worldview. By that stage, the story has taken on a life of its own. The money is irrelevant; the machine is already running on human conviction. Microsoft’s Threat Analysis Center has studied this network and concluded that its true strength is volume, not precision. The people running it know that most of their fabrications will fail. They do not care. They are playing a numbers game. If you flood the zone with enough outlandish videos and scandalous claims, some will inevitably slip through. That is not a bug in their system. It is the entire point. For every fake story that gets killed, a dozen more are waiting in the pipeline, each one cheap to produce and impossible to recall once released.
Perhaps the most unsettling part is what happens when the truth fights back. Mainstream media outlets have improved their fact-checking and often shut down manufactured rumors before they reach a nationwide audience. But debunking can be a double-edged sword. When a major news organization publishes a careful rebuttal, it also introduces the rumour to millions of people who had never heard it before. The correction may be accurate, but it carries the poison along with the antidote. Meanwhile, the official denial itself tells viewers that this story was serious enough to warrant a response, which for some is evidence that there must be something behind it. This is what researchers call the “liar’s dividend.” In an environment full of synthetic media, authentic evidence begins to lose credibility as well. If any video can be dismissed as a deepfake, then real footage becomes just one more file among many, no more trustworthy than the counterfeit. When that skepticism takes hold, the very idea of verification is weakened. Future corrections, no matter how rigorous, are met with a shrug. Glucksmann himself described the goal of these operations as creating a “halo of suspicion.” It is a powerful phrase because it captures something important: the suspicion lingers even after the lie is exposed. You can prove that a video was fake, but you cannot fully erase the question mark it leaves in a voter’s mind. Multiply that by dozens of operations run by separate networks, and the problem becomes nearly impossible to solve with fact-checking alone. Catching one burner account, one fake website, or even one entire network does not stop another network already working elsewhere. Three candidates attacked in a matter of weeks, by two independent operations, is just a sample of what that redundancy looks like over a full campaign.
France has not been passive in the face of this threat, at least on paper. Prime Minister Sébastien Lecornu credited a new detection system for automatically flagging all three campaigns as they emerged, a sign that the technical response is improving. A draft law addressing digital interference has already been presented to the Council of Ministers. The Paris prosecutor’s office has also opened formal investigations into the operations targeting Glucksmann, Attal, and Philippe. These are not meaningless gestures. They show that French institutions understand the gravity of the moment and are willing to use their tools. But the deeper question is whether speedier detection and new legal frameworks can truly outpace a pipeline built specifically to survive being caught. The architecture of Russian interference is designed with redundancy at its core. Even if one campaign is exposed within hours, the network can pivot to a new lie, a new candidate, a new social platform. Legal cases take years to conclude, by which time an election will have come and gone. The damage from a story does not wait for a judge’s verdict. Still, the shift in official thinking matters. Rather than chasing individual fabrications, French authorities are beginning to treat the system itself as the problem. They are asking how to make the environment more resilient, how to reduce the impact of the first fake, how to break the laundering chain before it reaches the mainstream. That is a meaningful evolution, but it remains to be seen whether it is enough.
What France is facing in 2027 is not a conventional propaganda campaign designed to convince voters of a single falsehood. It is something stranger and more corrosive. The goal is to create enough uncertainty that people stop believing anything at all. If a voter hears that one candidate has dementia, another has Parkinson’s, and a third is the puppet of a journalist who bribed the media, the cumulative effect is not belief in any one story. It is a vague, exhausting sense that the whole system is rotten. That feeling is dangerous because it does not require a specific lie to take root. It only requires a few seeds to be planted. Over time, the public may become less certain about what is true, less trusting of the institutions that are supposed to explain reality, and less willing to engage with the democratic process at all. That is precisely the outcome that may serve the interests of a foreign adversary trying to weaken Europe’s resolve on Ukraine and to elevate candidates who are sympathetic to Moscow. The concrete facts of the summer 2026 attacks are shocking: fake journalists, fake evidence, fake diseases, all traced back to a hostile intelligence service. But the deeper story is about trust, and trust is far harder to rebuild than a headline is to correct. As the campaign for the 2027 presidential election intensifies, with new candidates, new controversies, and new distractions, the question remains open. Can French institutions respond effectively enough to prevent this slow poisoning of the public sphere? The answer will depend not only on advanced detection systems and legal frameworks, but also on whether citizens themselves retain the will to demand evidence, to tolerate uncertainty, and to protect the democratic space from those who would prefer to see it dissolve into noise.