At first glance, nothing connects the drowning of dozens of young people off a Moroccan beach, a German schoolroom debate about the Holocaust, and a synthetic doctor on YouTube pushing prostate supplements. But read the last four editions of EDMO’s Weekly Pulse and the connection becomes uncomfortable and precise. In the online information space, all of these apparently unrelated stories are the same business model applied to different kinds of vulnerability. The model has two faces. One sells a future that will never happen: the open border, the hidden cure. The other sells a past that never existed: the heroic general, the golden empire, the innocent nation. The merchants of false hopes and the counterfeiters of historical memory are not different trades. They work opposite ends of the same human need, the need to believe that things could be otherwise. The raw material is not stupidity. It is desperation. Facing the gravest danger, people cling to whatever offers hope. A perfectly rational person who laughs at conspiracy theories on Monday can, upon receiving a diagnosis on Tuesday, fall into a rabbit hole of miracle cures. Almost anyone in a state of desperation can fall into these traps. That insight makes the Ceuta tragedy precise. The young Moroccans who swam around the breakwaters on 30 July were not credulous fools. They were people for whom the claim that Spain had opened a door was worth staking a life on, because the life being staked already felt unbearable. Maldita.es traced a trigger: a genuine Spanish Supreme Court ruling of 29 June, published on 8 July, establishing that summary returns cannot be applied to those who swim across, since they crossed no physical barrier, and that a return procedure must instead be opened within seventy-two hours. The ruling authorises nobody’s stay, but online it became “there will be no immediate expulsion,” and then “Ceuta is an open door.” Between 8 and 28 July the distortion circulated with little traction. Then, on 30 July, hundreds of posts carrying identical Arabic text appeared on Facebook, Instagram and TikTok. Some users published the same message ten times within minutes; one of the earliest amplifiers was a presenter on Moroccan state television with 1.3 million followers. Beneath one post, Moroccan users wrote: “I’m going to leave the country too.” Nearly a hundred people died. This is not misreading a legal text. This is hope reverse-engineering a legal text until it says what it must say. The fake oncologist works identically, offering cancer patients a version of reality in which the cure exists and is merely suppressed.
The third case, ultra-nationalism, targets a subtler desperation: the wish not to be ashamed. A nation that feels diminished wants a past worth belonging to. So Ratko Mladić, convicted of genocide, is mourned as a hero unjustly condemned by an anti-Serbian Western plot. So the AfD, having taken more than forty percent of the votes in Saxony-Anhalt, proposes reducing the weight of Holocaust and Nazi teaching in school curricula and abolishing the civic-education agency that funds visits to Holocaust memorials, while substituting a romanticised Bismarckian Kaiserreich for post-war remembrance. Calin Georgescu’s 2024 presidential campaign did the same in Romania with Dacian mythology, glorifying Codreanu and Antonescu for their “good deeds.” The product on sale is dignity, and the price is the hopeless surrender to a lie. Yet none of these operations emanates from nothing. Each takes something true and strips away the part that makes it inconvenient. The Spanish Supreme Court ruling was real; what circulated on social media omitted the return procedure. The German migration statistics debunked by GADMO were real; the influencer simply set 22,787 deportations against 1.48 million “new immigrants,” a figure that includes nearly 192,000 returning Germans and over 457,000 EU citizens, while Afghans and Syrians together accounted for only 5.2 percent of arrivals. The mebendazole patent held by Johns Hopkins University is real; the leap to “the cancer industry is hiding the cure” is not. Children with ADHD do show, on average, lower magnesium levels; the inference that magnesium treats ADHD does not follow, and the studies conflict. This matters enormously for defence. A detection regime built to catch fabrications will catch almost none of this. Decontextualisation leaves no forensic signature. The lie is not in any individual proposition but in the frame, and frames cannot be fingerprinted.
Where fabrication is needed, generative AI now supplies it at a scale that changes the nature of the problem. NORDIS partner TjekDet documented over a thousand AI-generated “doctors” on YouTube producing roughly 100,000 misleading videos in thirty languages. Fictional practitioners such as “Sofie Larsen” dispense advice to older Danish men; some channels impersonate the branding of renowned hospitals to manufacture credibility. IBERIFIER and Chequeado traced the migration equivalent of the same manipulation: AI videos of collapsing migrant-dominated London neighbourhoods, a fabricated Muslim mayoral candidate in Germany invented to provoke racist reactions, false AI images of Venezuelan migrants looting in Mexico. The historical variant may be the most vertiginous: Provereno Media found Facebook pages publishing AI-generated Holocaust victims, both faces and biographies invented, harvesting engagement from grief. The Auschwitz-Birkenau Museum called it falsification of history, and fact-checkers flagged the second-order danger explicitly: if some victims can be shown to be fake, Holocaust deniers acquire a tool for casting doubt on the real ones. That is the endgame of synthetic authority. It does not merely add false claims to the record; it corrodes the possibility of a record. EDMO BELUX’s media-literacy work on AI-generated “archive” images of concentration camps addresses precisely this: what becomes of historical evidence when images of the past can be conjured on demand? It is tempting to read all of this ideologically. But behind the AI doctors sit scammers selling remedies, plus influencers who spread identical lies simply because the content goes viral and platform monetisation converts virality into income. Behind xenophobic AI slops sit some politically motivated actors, but also content farms monetising sensationalism. Behind the invented Holocaust victims sit pages driven by likes, shares and advertising revenue. Ideology and greed are not rivals here; they are partners. The ideologue supplies the narrative, the profiteer supplies the distribution, and the platform supplies the incentive and collects a fee. Our judgement is blunt: platforms do not merely fail to react properly to crises, they fail on a daily basis, refusing to invest anything close to the resources needed to protect their own users, while profiting from demonstrably false viral content. EDMO Ireland’s analysis of the 2025 Irish presidential election quantifies it: 77 percent of flagged misinformation went unaddressed, and 87 percent of non-AI election disinformation was ignored entirely.
The single most important analytical insight across the last four editions of EDMO Weekly Pulse is this: disinformation narratives do not derive their power from intense campaigns during elections or crises. They derive it from the daily circulation of false content, like constant dripping wearing away the stone. Crises are exploited by disinformation to reinforce pre-existing narratives, not to create new ones. Watch the sequence after Ceuta. First migrants portrayed as zombies by far-right influencers. Then migrants eating cats, a narrative recycled straight from Trump’s 2024 Springfield claim. Then burning churches: a 2025 lightning strike on a church in Saint-Ours, Quebec, recirculated in Polish, French, English, Spanish and Vietnamese, as arson by a Muslim migrant, and further amplified in Poland by PiS MPs. Then Britain supposedly accepting 4,000 Moroccans from Ceuta, with invented murders attached. Then the phantom “second wave” of 15 August, built on footage of Fnideq from 31 July, promoted by Tommy Robinson in the UK, a Chega MP in Portugal, and an AfD politician in Germany. Millions of views each. Not one of these required the Ceuta crisis to exist. Ceuta merely provided an occasion to water a crop already planted. The cumulative message is that these “barely human beings” are incompatible with European societies and drastic measures are therefore welcome. The opinions themselves, zero tolerance for illegal migration, are entirely legitimate. The problem is that the ground on which opinions form has been poisoned. Foreign powers wait downstream. Spanish fact-checkers found no evidence that foreign interference triggered the Ceuta crossings, and the EEAS found no conclusive evidence that foreign state actors caused the crisis. What they found was exploitation afterwards: the Pravda network recycling a 2025 photograph of PSOE ministers at the Moroccan embassy as though taken during the crossings; pro-Israel accounts pushing #FreeCeuta to reframe Ceuta and Melilla as colonies; Russia-linked networks promoting the theory that Spain let it happen deliberately. Hostile actors, in other words, need originate nothing. They rather need a society already primed by daily dripping, and a platform economy that rewards whoever arrives with the most inflammatory framing.
This is why the memory question is a security question. As the 10 September Weekly Pulse observes, rewriting European history ultimately benefits powers outside Europe. Russia has much to gain from an AfD government, and it is no longer alone in cheering one on. What follows from this? EDMO’s first proposed remedy is unglamorous but correct. European fact-checkers, working with counterparts in neighbouring countries, in the Ceuta case African organisations working in Arabic, should systematically monitor platforms and local-language content, and issue early warnings that public authorities could convert into preventive communication. The expertise exists; the independence is guaranteed by EFCSN and IFCN standards. What is missing is money and people. Someone should have been watching Arabic-language Facebook feeds between 8 and 28 July, when the distortion was circulating without traction. The fuse was visible for three weeks before it was lit. EDMO is actively extending its network of fact-checkers to organisations in EU candidate, potential candidate and neighbourhood countries, despite its limited resources. This is a path that will be pursued further. But more is needed. Lack of adequate and predictable funding is one problem. The other is the absence of a dedicated EU operational structure with the authority to coordinate resilience-building activities and rapid response and escalation mechanisms connecting national operational capacities to EU-level coordination. Both problems have been acknowledged by the European Parliament in a resolution of 15 September 2026, endorsing the recommendations of the Special Committee on the European Democracy Shield. In July, another European Parliament resolution, pledging for a new strategy for media literacy and digital learning, called for strengthening EDMO activities on media literacy to support cooperation among Member States.
The second remedy, therefore, is about strengthening the institutional framework. In her State of the Union speech of 16 September, President von der Leyen expressly referred to the need to step up our “collective capacity to anticipate, detect and respond” to cognitive warfare. EDMO stands ready to pilot test the Stakeholder Platform that will operate within the newly established Centre for Democratic Resilience. But the upcoming EU Multiannual Financial Framework must also address the long-term sustainability of such efforts. The third remedy is harder. Regulating American tech giants is now more politically difficult and economically inconvenient than ever for the EU. But it remains the right thing to do, and “it can literally save lives,” a phrase written about fake doctors in the 17 September Weekly Pulse, but earned just as fully on a beach in Ceuta. Today the targets are migrants, cancer patients, and the murdered dead who cannot object to being forgotten. But what about tomorrow? We are all, at some point, someone who needs hope, someone who wants to belong to a story worth telling, someone who is tired and afraid and looking for an explanation. The systems described here are not abstract threats. They are built to meet us in our most human moments, and they are built to fail us there. The only way to resist is not to become harder or more cynical, but to learn to recognise the shape of the lie beneath its comforting surface: the false cure, the false border, the false past, the false hero. We can choose to hold on to the messy, inconvenient, difficult truth instead. It is less comfortable, but it is all we have, and it does not require anyone to die.