At the Heart of the Scandal: How I Examined the Ibiza Video for Authenticity

It began with an email. Der Spiegel asked whether I could analyze video and audio in a matter that no one would discuss over the phone. I knew the magazine, I knew the weight that sits inside such a carefully worded request, and I said yes without sensing what I was walking into. A few days later two reporters were standing in my Munich office, with a camera team, with grave faces, with a data carrier whose contents knocked me sideways in the first few minutes. I saw what was on it. And I knew at once, this has potential. Not the pleasant kind of potential. The other kind.

What I signed that day was a non-disclosure agreement, an NDA, and I have honored it to this day. I will not let slip anything here that falls under that agreement. What I can tell is what has long been public, what stood in the newspaper, what can be read in a book, what a television documentary shows with me sitting in front of the camera, and what the encyclopedia entry on the affair records, naming me by name. And I can tell what it feels like to sit alone for weeks with material that ends up dismantling a government.

Seven years have passed since then. Long enough to write about it soberly. Short enough that one question still will not leave my head, and I want to close with it, because it is more honest than any answer I could give. Let us start at the front, with the material itself, because without the material you understand neither the uproar nor the work that followed.

What the Material Actually Showed

The recordings were made secretly on 24 July 2017 in a villa on the Spanish island of Ibiza. This villa had been rented by persons unknown for the express purpose, from 22 to 25 July, a detail worth remembering, because it shows how deliberately the whole thing was set up. Nobody rents a finca for three days, installs cameras and lures two senior politicians into it without a plan and without means. The total material ran to over 20 hours, video from several perspectives and audio taken together. The audio track of the conversation alone ran over 7 hours. Over 7 hours that I listened through night after night, in one stretch, then again from the start, then in fragments, until I recognized every vocal color in my sleep.

The footage shows two men who at the time stood at the top of power in Austria. Heinz-Christian Strache, then Vice-Chancellor and federal party leader of the FPÖ. And Johann Gudenus, his close confidant, who took on the role of interpreter at times because he spoke Russian. Across from them sits a woman who passes herself off as the niece of a Russian oligarch, the niece of the oil and gas magnate Igor Makarow, with a supposed Latvian passport and the claim that she wanted to invest several hundred million euros in Austria. She was none of that. She was a hired decoy. Makarow himself later stated publicly that he was the only child in his family and had no nieces at all. The rich Russian woman was a staging, and the two men walked straight into it.

What Strache produced over those hours is the real explosive charge. He held out the prospect of state contracts to the supposed investor, all the contracts that until then had gone to a large domestic construction group and its prominent owner, provided she founded a construction firm of her own. He toyed with the idea of privatizing a channel of the public broadcaster and handing the water supply, down to a blocking minority held by the state, into private hands. Above all, though, he spoke, again and again that evening, of a takeover of the Kronen Zeitung, the country's largest daily. If it belonged to her, he said, one could speak quite openly, one could build up a few people and get rid of a few others, and if the paper suddenly backed the Freedom Party two or three weeks before the election, a vote share of 34 percent was within reach.

And he described how the money should flow, without a trace. A donation to a charitable association, with the fine statute of making Austria more economically sound, with three lawyers on its board, so as to bypass the court of audit and the laws on party financing. In between, almost defiantly, the sentence that everything had to be lawful and legal. That mixture is the genuinely disturbing part. A man who in the same breath sketches the circumvention of the law and insists on its observance. He called himself the Red Bull Brother from Austria that evening, and the journalists who would later expose him he called the worst the planet has to offer. The whole evening he later termed, in his own words, a drunken affair.

All of that was published on 17 May 2019, at exactly 6 p.m., by the Süddeutsche Zeitung and Der Spiegel, simultaneously, in the form of six-minute excerpts. Both houses stressed that they had paid nothing for the material. What went unpublished were the passages in which Strache disparaged the private lives of other politicians, which he dismissed at his resignation as unverified dirty rumors. Before any of it could be printed, a single question had to be answered beyond doubt. Is this real. That is exactly where I came in.

The Nights I Spent in the Wrong Movie

How the material reached the journalists is a thriller in itself, and it is by now public. They had first been shown only a 15-minute excerpt and then made to wait for months. At the first contact they were required to wear special glasses so they could not secretly film. The actual handover ran, as one of the reporters later put it, like something out of a bad film. They were steered for hours to a remote location, made to wait at a petrol station and finally led into an abandoned hotel. Only afterward did they have the complete video, and only then could my work begin.

In the days after the reporters' visit I switched into a mode I know from big cases and that my family appreciates less than I do. Survival mode, you might call it, except that it was not about survival but about concentration without pause. From then on I was practically only ever in the office, mostly at night, because at night nobody calls and the phone stays silent. The desk lamp, the headphones, the material. Nothing more was needed, and nothing more was allowed in.

First the audio. Over 7 hours of conversation that you do not simply play but dissect. You do not listen to the content, at least not first. You listen to everything else. To the room's baseline hiss, to the clink of glasses, to the way a voice sits in the reverberation of the walls, to the tiny sounds a person makes when shifting on a sofa. For the authenticity examiner the content is almost a distraction. Anyone who lets himself be gripped by what is said no longer hears how it is said, and the how is precisely my subject. A forensic examiner who lets himself be swept up in the drama is a poor forensic examiner. I once described in an earlier text how treacherously cognitive bias works in our field, how quickly one finds what one wants to find. In this case the temptation was greater than ever, because the material was politically white-hot. All the more disciplined did I have to keep the one question apart from all the others.

Concretely that means a rigor that looks almost absurd from the outside. I write down what I expect to find before I search, so that afterward I can check whether I really found it or merely expected it. I work against my own thesis, not for it. When I am convinced the video is real, that is the moment I search hardest for signs of forgery, because the most dangerous bias is the one that feels right. Over 7 hours of audio and the matching picture forgive no convenience. You hear a passage ten times, you slow it down, you set the spectrogram beside it, and only when nothing tips over do you tick it off and move to the next.

Then the picture. Frame by frame, second by second, matching what the camera shows against what is physically possible. At some point during those nights a feeling settled in that I had not known before. I was sitting in the wrong movie. Scenes were playing out as if from a badly invented political thriller, an oligarch's niece, a villa, hidden cameras, a trap, and I, the sober examiner, in the middle of it, with nothing but a lamp and a pair of headphones. The absurd thing is that years later this material genuinely became a film, several in fact, and I got to watch a production team turn my sleepless, glamour-free reality into high gloss. But more on that later.

What occupied me most during those nights was not the question of whether the material was real. That question I could answer, that is what I was there for. It was the question of how well the whole thing had been done. The placement of the cameras, the handling of light, the patience with which someone here had captured a conversation over hours without the two men growing suspicious. Whoever set this up understood their craft. That is an observation, not praise, and I will come back to it at the end.

What Authentication Really Means

Here I have to explain briefly what a forensic authenticity examination of moving images actually is, because the notions of it are usually wrong. There is no button you press and no software that spits out real or fake after a minute. It is laborious, multilayered detail work, and it falls roughly into three visible layers and one invisible layer that interlock.

The first layer is the visible one, the spatial reconstruction. I matched the perspectives, the light sources and the cast shadows in the video meticulously against the real dimensions and furnishings of the finca. Where does the camera stand, at what angle, how does the light of a lamp fall on a cheek, where does the shadow of a glass lie on the table, does all of it fit a single, physically consistent space. A real scene is a closed physical system. Every light produces exactly the shadows its position dictates, every surface throws back exactly the reflection its material permits. When a person is cut into such a scene afterward, by green screen or by computer graphics, that closure breaks somewhere. A shadow points the wrong way, a reflection is missing, an edge sits a hair's breadth off, a patch of skin is lit a shade too cleanly. None of that was to be found here. Several camera perspectives make the matter even more hopeless for a forger, because he would have to keep the same manipulation consistent across every perspective, and that practically never succeeds. The team at the Fraunhofer Institute for Secure Information Technology, with which the examination ran in cooperation, went the same way in its own fashion and compared, for instance, the stone patterns on the house walls with the advertising photos of the rented villa. If the masonry in the video matches the masonry in the listings, then the claimed location holds. There too, everything fits together.

The second layer is the audible one, the audio and voice analysis. An audio track running over hours has a signature. The baseline hiss stays constant, the room reverberation stays constant, the device doing the recording leaves a steady fingerprint in the frequency picture. When someone inserts, cuts out or rearranges sentences afterward, breaks appear. A splice at which the hiss flips for a fraction of a second, a pitch that does not match the breathing, a reverberation that suddenly sounds like a different room, a transition too clean to be real. I checked the tracks for frequency constancy and searched for exactly those breaks. There were none. Nobody put sentences into Strache's mouth that he did not say, and nobody reversed the order to turn harmless words into dangerous ones.

The third layer is the one that rarely sits in a single pair of hands together with the first two, and that was precisely my advantage here. I do not work only digitally, I also work with the methods of bodily measurement, with proportions, with movement mechanics, with the analysis of lip movements in relation to the spoken word. With that you test whether the persons in the picture really are the ones speaking, and whether they really speak what you hear them say. That is the bolt against the suspicion that arises first with every explosive video, all the more today, in the age of artificially generated faces, namely deepfake or disguised doubles. A deepfake betrays itself in the mechanics, in transitions between expressions, in the fine delay between sound and lip, in proportions that do not stay stable over time. A disguised double betrays itself in body dimensions, in gait, in posture. I examined the participants anthropometrically and biomechanically, measure against measure, movement against movement. Nothing here was forged. They were the men you saw, and they said what you heard them say.

The fact that these three approaches lie in a single pair of hands with me is no accident, and in this case it was decisive. A pure data forensic examiner sees the metadata but not necessarily whether a body moves mechanically correctly. A pure image examiner sees the scene but not necessarily the trace in the container format. Whoever can do both and on top of that masters the measurement of the human body closes the gaps through which a forgery would otherwise slip. Models and checklists do not replace that. It is the years of work on real material that train the eye, and exactly this experience was here the difference between a cautious maybe and a clear finding.

Over all of it lies the invisible layer, the metadata. Timestamps, file formats, sampling rates, codec signatures, the traces that every cut and every re-encoding leaves in the innermost part of a file. They tell the technical life story of a video, and they betray manipulation often where the eye is long since convinced. A subsequently edited clip almost always carries the scars of the editing program, a second compression layer, an altered structure in the container format. This layer too fit into a coherent picture. Four layers, one answer. At the end of this work stood a finding as unambiguous as a finding can be. The head of IT forensics at the Fraunhofer Institute later summed up the public result in a sentence that spoke straight from my heart, that they had come to the conclusion that all of it was authentic, that they had been unable to detect anything unusual.

There is a thought I owe to honesty, one that makes this case technically special. It was at once easy and hard. Easy, because no manipulation had even been attempted, for the makers wanted the material to be real and to look real, that was the whole point of the trap. A forgery would have defeated the purpose. Hard nonetheless, because at this magnitude every single statement had to be court-proof. I was not allowed to trust that nothing would turn out to be forged, I had to rule it out positively, layer by layer, and in a way that still holds when an opposing expert pulls every line apart months later. Precisely this asymmetry between a plausible assumption and a defensible finding is the difference between an impression and an expert opinion.

The Sentence an Honest Forensic Examiner Has to Say

And now comes the part most people skip, although it is the most important. Real and unmanipulated means: the file was not forged. It does not mean that a six-minute cut out of over 7 hours of conversation conveys the sense of the words correctly. And it certainly does not mean who set the trap and why.

Those are three different questions, and you have to keep them cleanly apart, otherwise you make yourself vulnerable as an expert. Authenticity is the one. Context is the second. Authorship is the third. The first I could answer, with great certainty. About the second anyone can argue who holds a six-minute excerpt against over 7 hours of raw conversation, and Strache did exactly that, he spoke of statements torn out of context. That is the standard defense, and in principle it is even legitimate, because a cut can shift meaning without falsifying a single word. Only it rested on nothing here, because the Austrian journalist who was allowed to see the complete material before publication later summarized before the parliamentary committee that the content supported the corruption allegations, and the newsrooms had even met the supposed oligarch's niece in person before publishing. The third question, the one about the makers and their motives, lay entirely outside my brief. A forensic examiner who blends these three layers does not deliver an expert opinion, he delivers an opinion, and opinions were plentiful in this case.

An example makes the difference tangible. Imagine someone cuts a question out of a long conversation and places beside it an answer that originally belonged to a quite different question. Not a single word is forged, every syllable is real, and yet the splice lies. A pure authenticity check at the level of the data alone might not even notice, because the audio snippets themselves are unchanged. That is why I examine not only the file but also the continuity, the running hiss, the unbroken flow of conversation, the logic of the transitions. This continuity was precisely intact here, across the full length. Whoever wants to claim more, namely that the sense of the words was captured correctly, steps onto the field of interpretation, and that is the task of journalists, lawyers and in the end the public, not of the forensic examiner.

There is a second finding that does not come from me but points in exactly the same direction, and I mention it here for that reason. Strache and Gudenus later suggested that knockout drops might have been administered to them that evening, that this would explain their behavior. Two of the most renowned forensic physicians in the German-speaking world, Michael Tsokos, head of the Institute of Legal Medicine at Berlin's Charité, and Sven Hartwig, head of the Charité's forensic toxicology, examined this on the basis of the public video and called it simply impossible. Had the two been given a common knockout agent, increasing fatigue would have shown over the hours, drowsiness, sleepiness, in the end perhaps clouding of consciousness. Instead, the experts wrote, body language and conversational behavior were unremarkable, the gestures content-related, word choice and grammar even sophisticated, and that over several hours, including complex matters that Gudenus translated into Russian in between. In plain terms, the man was in full possession of his faculties. The authenticity finding and the toxicological finding together yield a clear picture. The material is real, and the man you see in it meant what he said. Whether that was politically wise or morally defensible is not my department. Whether it was real is, and it was real.

Submission Day and the Earthquake

On the day of submission I delivered what I had found. No forgery. Not remotely possible. All the data handed to me and the expert opinion I had drawn up were deleted upon handover, as the agreement I was bound by demanded, and so it was done. What remained with me was nothing but the memory of weeks in front of the screen.

Then everything happened very fast. At 6 p.m. on 17 May the material went online. Even in the run-up a German satirist had made allusions to the contents at an awards ceremony in April, which at the time passed as an exaggerated joke and which accelerated the publication that had been planned anyway. At midday on 18 May, less than a day after publication, Strache resigned as Vice-Chancellor and party leader, in a statement that called the video a targeted political assassination and spoke of a smear campaign. On the evening of that 18 May, Chancellor Sebastian Kurz declared the end of the coalition, called the way the recording had come about contemptible but the content telling, and proposed new elections.

What followed was a lesson in accelerated state collapse. Kurz demanded, for a continuation of the coalition, the removal of the Freedom Party's interior minister, the FPÖ refused, and its ministers resigned as a bloc. On 27 May the National Council passed a vote of no confidence in the entire federal government, for the first time in the history of the republic. A day later the members of government were removed from office, and a caretaker cabinet under the jurist Brigitte Bierlein took over. New elections were held on 29 September 2019, the People's Party gained sharply, the Freedom Party lost, and a coalition with the Greens emerged. In January 2020 a long parliamentary committee of inquiry took up its work, which would occupy half of political Vienna. A few days after publication, the Austrian government was in ruins.

The echo reached far beyond Austria. In Germany senior politicians commented on the case, in France the leader of a party allied with the FPÖ wondered aloud about the timing of the release shortly before the election. The later committee of inquiry worked its way for months through donation constructions, appointments and document deliveries, up to the question of who knew what and when. None of that had anything to do with my work anymore. My part was a tiny, technical foundation right at the start, and on this foundation a political earthquake built itself up, step by step.

I sat in Munich and watched a file I had taken apart night after night set a country in motion. It is a strange feeling to stand at such a small, technical point in such a large story. I decided nothing, I published nothing, I answered only a single question, whether the thing is real. But without that answer no one could have printed. The authenticity check is the quiet point at which it is decided whether a suspicion may become a publication. It stands at the start of the causal chain, right at the bottom, invisible, and it carries the full weight of everything that comes after. Whoever errs here either condemns an innocent or lets a scandal escape as a forgery. Both would be a catastrophe. Hence those weeks, hence the nights, hence the four layers.

How the Case Became a Film

The strangest thing about this story is what happened to it afterward. The Süddeutsche Zeitung granted a glimpse into its own workshop a few weeks later and explained on 28 June 2019 how the Strache video was examined for authenticity. In that piece it stands in black and white, the data forensic examiner George A. Rauscher established that the video was not a forgery. That is not my own account of myself, that is my client naming me as the forensic examiner who established authenticity.

The two investigative journalists of the Süddeutsche, Bastian Obermayer and Frederik Obermaier, the same who had earlier evaluated the Panama Papers, wrote a book about the case. Die Ibiza-Affäre, Innenansichten eines Skandals, published in 2019 by Kiepenheuer und Witsch. Two years later Sky made two productions out of it, a four-part fictional series and a ninety-minute documentary titled Das Ibiza-Video, ein journalistischer Krimi, both aired in 2021 and repeatedly honored. In that documentary I am one of the interview partners, alongside the two journalists, alongside a Viennese colleague, alongside a former bodyguard of Strache, alongside connoisseurs of Austrian politics. And the encyclopedia entry on the Ibiza affair states to this day that the authenticity of the video was examined by reporters of the participating media and the Munich forensic examiner George A. Rauscher in cooperation with the Fraunhofer Institute.

Here is the small irony I cannot resist. I lived the unglamorous version, nights, headphones, cold coffee, no champagne, no sushi, no villa with a sea view. And then I sat there and watched a high-gloss production turn precisely those weeks into prestige television, with actors, with music, with dramatic arcs that my real life in those days sorely lacked. To become part of a story whose authenticity you examined yourself is a strange feeling. You stand in the closing credits of reality long before you stand in the closing credits of a film. And you know that the sober truth about the work you did appears in no dramaturgy, because it consists of patience, and patience films badly.

And yet something about the documentary reconciled me. At its core it is a plea for clean journalism, for the patient checking of facts in a time when every claim finds its way onto the net at once and unchecked. The director put it roughly like this, that in times of permanent news flood and dubious sources many people are left asking whom they can still trust at all. That is exactly where my profession sits. The authenticity check is the inconspicuous threshold between a rumor and a documented fact, and whether that threshold holds decides whether a society can stand on a shared factual ground or sinks into mutual suspicion.

The Question That Stayed With Me

And with that I arrive at what stays with me to this day. My work answered one question, whether the material is real, and the answer was a clear yes. It never answered another question, and could not answer it, because it is not a forensic question. Who built this, and why, and why precisely at that moment.

I conjure nothing up here, I tell no intelligence-service story, my reputation is too dear to me for that. I name only what is documented and held on the court record. The video was made by a Viennese lawyer and a detective, and over them I will pass no judgment, for as reprehensible as the method was, in the end they too stand there as driven men, hunted, charged, entangled in something that grew over their heads. An Austrian higher regional court called the manner of obtaining the information dishonest to a particular degree and unlawful in several respects. The supreme court nonetheless lifted the publication ban, because the material was an extraordinarily large contribution to a debate of public interest. Both are true, and both at once. A method can be unlawful and the result still of overriding public value.

Three things give me pause to this day, and none of them is a conspiracy, each of them stands in the files. One of the makers stated that he had spent, directly and indirectly, at least 500,000 euros on the production. That is a great deal of money for an idea supposedly born of outrage over alleged abuses. The material surfaced, of all moments, a few days before a European election, a timing that served the political opponents of the Freedom Party to the maximum. And a large German newspaper later reported there were indications that a third group had paid to channel the video to the two newsrooms. Three observations, no verdict.

I say no more, because I know no more. I am the man who held the file in his hands, and I can say with certainty only one thing, it was real. The rest, the question of purpose and of timing, is no case for a frequency diagram and none for a cast shadow. It is the question that stays open, and it stays open because the people who could answer it are silent, and the people who have spoken have not said everything. Sometimes the most honest result of an investigation is the clean finding, followed by the admission that behind it lies a door you cannot open. Behind that door stands my question. It will stay with me as long as I think about this case, and I will, every time I hear that a video is too good to be true, or too terrible.

This article is a revised and expanded new version of my original text on the Ibiza video from 2019. The verifiable facts rest on publicly accessible sources, the personal accounts on my own memory of the case. Everything covered by my non-disclosure agreement stays where it belongs.

References

• Wikipedia. Ibiza-Affäre. Retrieved 02.06.2026. https://de.wikipedia.org/wiki/Ibiza-Aff%C3%A4re
• Obermayer, B., and Obermaier, F. (2019). Die Ibiza-Affäre. Innenansichten eines Skandals. Cologne: Kiepenheuer und Witsch. ISBN 978-3-462-05407-1.
• Süddeutsche Zeitung (2019, June 28). Wie das Strache-Video auf Echtheit geprüft wurde. Video by Annika Sehn. https://www.sueddeutsche.de/politik/digitale-forensik-wie-das-strache-video-auf-echtheit-geprueft-wurde-1.4503178
• science.ORF.at (2019, May 20). Wie das Ibiza-Video überprüft wurde. https://science.orf.at/v2/stories/2982565/
• Steinebach, M. (2019). Echtheitsprüfung von Videos mit Beispielen aus der Ibiza-Affäre. Fraunhofer Institute for Secure Information Technology SIT, Anwendertag IT-Forensik. https://www.sit.fraunhofer.de/
• Neue Zürcher Zeitung (2019, May 21). Strache-Video, ein Skandal mit geschickten Regisseuren. https://www.nzz.ch/feuilleton/medien/strache-video-ein-skandal-mit-geschickten-regisseuren-ld.1483270
• news.ORF.at (2019, September 26). Gutachten zu Ibiza-Video, keine K.-o.-Tropfen im Spiel. https://orf.at/stories/3138642/
• Sky Deutschland, via Presseportal (2021, September 8). Die Dokumentation zur fiktionalen Serie Die Ibiza Affäre, Das Ibiza-Video, ein journalistischer Krimi. https://www.presseportal.de/pm/33221/5014677
• Sky Originals. Die Ibiza Affäre and Das Ibiza-Video, ein journalistischer Krimi. https://mediencenter.sky.at/
• Rauscher, G. A. (2019). IbizaGate, I examined the StracheVideo. https://rauscher.xyz/ibizagate/