Photo of Lee Jae Myung with Trump at G7 is real, and the image used to accuse manipulation was generated by AI
The 2026 G7 summit was still making waves when a controversy erupted on social media involving South Korean President Lee Jae Myung and U.S. President Donald Trump. A photo published by Lee himself on X on June 18 showed the two leaders together during a dinner at the event held in France. It did not take long before users started claiming the image was a fake, supposedly created from an old photo of Trump with imprisoned former president Yoon Suk Yeol. The problem is that the story flipped completely upside down once fact-checkers got involved.
The photo of Lee with Trump is real and authenticated. A spokesperson for the South Korean presidential office confirmed to AFP on June 19 that the image is authentic and was not generated by artificial intelligence. Meanwhile, the image circulating as the supposed original photo of Yoon with Trump was analyzed by the Hive Moderation tool and showed a 96.9% probability of containing AI-generated content or deepfake. In other words, whoever tried to prove a photo was fake ended up using exactly that, an image fabricated by artificial intelligence, to support their argument. It is the kind of plot twist that perfectly sums up the current landscape of disinformation through image manipulation around the world 🌐
What happened during the G7 summit in France
The 2026 G7 summit took place between June 15 and 17 in the French city of Evian, with the luxurious Hotel Royal serving as its backdrop. During the event, world leaders gathered to discuss highly relevant global issues, including an agreement signed by Trump aimed at ending the conflict in the Middle East. South Korean President Lee Jae Myung attended the summit as part of the country’s diplomatic agenda, and after bilateral meetings with various leaders, he posted a photo alongside Donald Trump on X. The image captured a moment during the official G7 dinner held on June 16 at the Michelin-starred restaurant Les Fresques, located inside the Hotel Royal itself.
In addition to the photo with Trump, Lee also shared an image of a pen he said he received as a gift from the American president. Both images were featured in reports by South Korean media outlets such as MBN and the Korea Herald, reinforcing the legitimacy of the record. Lee stated that the two leaders discussed South Korea’s longstanding standoff with North Korea, its nuclear-armed neighbor. According to the South Korean president, he told Trump that Russian assistance had rendered the sanctions imposed on Pyongyang ineffective. There was also speculation that the Trump administration would turn its diplomatic focus to the North Korean issue after the summit.
The photo quickly gained traction on social media, but along with the positive engagement came the first waves of skepticism, especially among political groups that had no interest in seeing the two leaders’ rapport being celebrated. It was in this already heated environment that claims began circulating that the photo was actually an image manipulation produced with the help of AI.
The false claim and how it spread
The theory that gained momentum said that the image of Lee Jae Myung alongside Trump had been created from a supposedly original photo of former President Yoon Suk Yeol with the American leader. The idea was that Yoon’s face had been swapped with Lee’s through deepfake techniques. A post in Korean on Threads, shared on June 18, 2026, claimed that the photo of Lee with Trump was a skillfully manipulated fake composite image. The post included the two images side by side, as if that were sufficient proof of the fraud, and featured a screenshot that allegedly showed the photo of Yoon with Trump had originally been published on X on April 30, 2023.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
The narrative spread at impressive speed, being shared in both Korean and English by various profiles across social media. What few people noticed in the heat of the moment was that nobody had actually verified where this supposed original photo of Yoon with Trump came from. The image circulating as the reference had no confirmed origin, did not appear in reliable news archives, and had not been published by any official source. But that did not stop it from being shared thousands of times as though it were a legitimate historical document. This is exactly the dynamic that makes disinformation so dangerous: the speed of sharing far outpaces the speed of verification.
The evidence confirming the authenticity of the Lee-Trump photo
The fact-check conducted by AFP and specialized tools revealed multiple layers of evidence confirming that the photo published by Lee Jae Myung is genuine. First, the analysis of the image by Hive Moderation’s AI detection tool found no indication that the photo had been artificially generated. The image passed authenticity tests without raising any red flags.
Beyond that, visual elements in the photo are consistent with other records from the same event. The ties that Lee and Trump are wearing in the image match exactly those captured in photos distributed by AFP before the official G7 dinner on June 16. The background setting also matches the decor of the Les Fresques restaurant at the Hotel Royal in Evian, the venue where the summit was held. To top it off, a photo from the same meeting captured from a different angle and published by the Korea Times confirms the interaction between the two leaders in that same setting and on that same occasion.
A spokesperson for the South Korean presidential office officially reaffirmed to AFP on June 19 that the photo is authentic. All of this combined forms a robust body of evidence that leaves no room for the manipulation theory.
The photo of Yoon with Trump never existed
If the photo of Lee with Trump is demonstrably real, the image of Yoon Suk Yeol with the American president that was used as supposed proof of fraud tells a very different story. The analysis of this second image by the same Hive Moderation tool revealed it had a 96.9% probability of containing AI-generated content or deepfake. In practical terms, the photo presented as the legitimate original was actually a synthetic creation.
The fact-checks went even further. A search for posts made by Yoon Suk Yeol’s official account on X on April 30, 2023, the date the false posts claimed was the original publication date of the photo with Trump, turned up no relevant results. What Yoon actually posted on that date was a note of gratitude for the warm reception he received from then-President Joe Biden and First Lady Jill Biden during a state visit to Washington. In the post, Yoon referenced a video from Biden’s official account highlighting the shared beliefs that form the foundation of the U.S.-South Korea alliance.
That meeting between Yoon and Biden took place in the context of a historic state dinner at the White House on April 26, 2023, organized to celebrate the 70th anniversary of the alliance between the two countries. The event received extensive press coverage, with videos showing Yoon singing excerpts of American Pie by Don McLean. A separate search of Trump’s social media accounts also confirmed that he did not post any photo with Yoon on that date. The official archive accounts of the 45th President of the United States and his first administration had not been updated since January 2021.
When AI enters the disinformation game
This episode completely flips the original narrative. The accusation that Lee had used image manipulation to forge a meeting with Trump ended up being propped up by an image that is itself fake. It is a paradox that would be almost comical if it were not so revealing about the current state of digital information. Whoever built the fraud argument used as their central piece of evidence the very thing they were trying to call out: synthetic content generated by AI. This is not an isolated mistake. It is a pattern that keeps repeating across different countries and political contexts, where realistic image generation technology is now being used both to attack and defend narratives.
What this episode makes clear is that AI is no longer just a tool for productivity or creativity. It has become an active instrument in the disinformation arena, capable of generating content convincing enough to fool untrained eyes and fuel real political disputes. The sophistication of image generation models has reached a point where distinguishing real from synthetic requires specific technical tools, not just the critical thinking of an average observer. This places enormous responsibility on platforms, news organizations, and also on the users themselves who share content without prior verification 📸
It is worth noting that AFP had previously debunked other misleading claims targeting Lee Jae Myung’s relationships with other world leaders, indicating that this kind of visual disinformation attack is not an isolated case but part of a broader pattern of digital campaigns against the South Korean president.
Why this case matters beyond politics
It is easy to look at this story as just another political fight on social media, but what it reveals goes far beyond that. Image manipulation with the help of AI has reached a level of quality that allows convincing false evidence to be created in a matter of minutes, and this has direct implications for public trust in visual records of any kind. Photographs have always been treated as evidence, as objective records of reality. When that premise starts to erode, the impact extends well beyond social media and reaches the perception of historical events, judicial decisions, journalistic coverage, and diplomatic narratives like those involving the G7 summit.
In the case involving Lee Jae Myung and Donald Trump, the most immediate damage was to the South Korean president’s reputation, as he had to deal with a very serious accusation at the exact moment he was trying to project an image of international leadership. But the long-term damage is more subtle and more profound: each episode like this contributes to an environment where people begin to distrust everything, including real images. When distrust becomes generalized, it gets increasingly difficult to tell what is real, and that creates fertile ground for any claim, true or false, to be dismissed with a simple label of fake.
The role of detection tools and fact-checking
One of the most relevant aspects of this case is how it demonstrates the growing importance of tools designed to detect AI-generated content. Without platforms like Hive Moderation, which can analyze images and assign probabilities of synthetic origin, it would be infinitely harder to dismantle narratives like the one that tried to discredit the photo of Lee with Trump. The technology that creates deepfakes and false images is the same that, on the other side, develops mechanisms to identify them. It is a constant race between creation and detection, and for now the verification tools can still keep up, but nothing guarantees that will always be the case.
The work of fact-checking agencies like AFP also proved essential in this episode. The fact-check was not limited to analyzing pixels and metadata. It involved researching historical records, cross-referencing dates, consulting official sources, and contextually analyzing multiple images from the same event. This kind of multidimensional investigation is what separates a serious fact-check from a simple opinion posted on social media.
The solution to the visual disinformation problem certainly does not come down to any single technology or regulation. What AI detection specialists and disinformation researchers have been pointing out is the need for a combination of technical verification tools, media literacy education for the general public, and greater accountability from platforms that allow synthetic content to spread without any labeling or warning. The G7 summit episode is a reminder that this debate is neither abstract nor futuristic. It is happening right now, with real consequences for real leaders, at real events. And AI-generated image manipulation is one of the key pieces of a puzzle that society is still trying to put together 🧩
