AI-Generated GTA 6 'Leaks' Fool Millions as Fake Gameplay Videos Go Viral

Our Enshittified Internet Can't Handle The Lack Of New GTA 6 Info

Rockstar's long-awaited blockbuster is proving an irresistible target for viral slop merchants "GTA 6 Keeps Trending Despite Zero New Info," is a headline I wrote back in 2021. It's even more true today, except that the spread of generative AI tools now makes it easier than ever to trick people into thinking there is new info about Rockstar's delayed open-world sequel. With just a few clicks, you can spawn convincing fake footage and send it masquerading around the internet until it becomes embedded in the algorithmic machinery and goes viral in one way or another. The stakes are low, but it's a perfect example of how being online keeps getting worse. "GTA 6 ALERT â€" EXTREMELY SERIOUS SITUATION" the verified X account Zap Actu GTA6 posted on Tuesday. It was re-sharing allegedly leaked footage of the upcoming game found on TikTok and hyping it up due to the "sheer level of realism" including "shockingly lifelike" NPCs and a "dynamic weather system unlike anything seen in GTA V." The footage is 100 percent fake, but that didn't stop the post from generating over 4.2 million impressions as it zipped around Elon Musk's third-rate Reddit knock-off. It illustrates the collision of three distinct but related developments: industrial-grade AI slop, algorithms designed to make that slop go viral, and the enshitification of a major social media platform controlled by a single, unhinged, soon-to-be trillionaire. None of this felt great four years ago, but my god, how quickly did it all get much, much worse. The problem's not isolated to X, either. Earlier this year, YouTuber Jeffery Phillips made it his mission to convince Google that GTA 6 had a twerk button. He did this simply by posting over and over again on X and Reddit about how the twerk button was real. The feedback loop eventually succeeded at tricking Google's internet-scrapping AI Overview search tool. "I think I spent a total of three months doing this â€~challenge,’ he told Polygon last month. "I started it in July and it was finished late September. I have no idea how much time I spent posting but it wasn’t an insane amount. Some days I’d just post one or two things.†It's the most brazen example of Google's new AI search tool getting suckered into such a high-profile and obvious mistake, but there are tons of smaller mistakes it makes that no one ever bothers to talk about or make a big deal out of, like what happened to Steam's merchant sim Trash Goblin, a game about polishing up junk and selling it. Its developers found that Google kept giving players incorrect information about the game and how it works. “As a programmer, if I type something into a computer, I expect it to give out the same output over and over," one of its developers told Game File. "Google results give something different every time, in some of these cases. They’re not learning better each time. Sometimes you get the right thing, and the next time you get something wrong again." But Google isn't just the gateway to the written internet. It also controls YouTube. Earlier this month, it sent thousands of people to a fake GTA 6 trailer that it happened to recommend to people via notification on their phones. Open "Grand Theft Auto VI - Final Trailer (2026) Rockstar Games" and you'll be treated to an elaborate AI-made fan concept trailer for what someone thinks the game could be like. It has over 600,000 views, 18,000 downvotes, and 2,000 comments. "No way everyone sent here by google lmao," reads the top one with over 7,000 likes. "Google fucking trolled us lmao," reads the second. "Screw that Google notification," reads the third. My favorite comment summed up this unconventional rickroll like this: "Ay no way we getting trolled by Google for a gta6 AI trailer before GTA 6." For much of its recent history, the internet has been a negotiation between humans and algorithms. Humans make and share stuff. The algorithms help organize and sort it. A few people get very rich. Hopefully, the rest of us come away slightly better informed or at least entertained. Generative AI threatens to blow up that entire deal. It was never very good to begin with, but what's coming to replace it is going to be so much worse.

GTA 6 fans grab their spare pitchforks and torches as AI "leak" gets out of hand, turning 8 million views into an angry mob: "I wouldn't have thought it would go this far"

One GTA 6 "news" account is suddenly finding itself in enemy territory as Rockstar fans, its former compatriots, realize its exclusive game footage is actually AI. Though the Zap Actu GTA6 account has already removed its generated videos from Twitter - along with several other posts referencing them - following a report from IGN, those who have been duped are still too enraged to be satisfied. The purported GTA 6 "leak" gained 8 million views in only a day, despite some fans warning each other that any video coming out of Zap Actu GTA6's account was worthless. Undeterred at first, Zap Actu GTA6 boasted that "the situation is escalating fast" and "MASSIVE LEAK IMMINENT." But as more people uncovered the truth, Zap Actu GTA6 discovered the unholy powers of a GTA player scorned. "I'm sorry if I raised false hopes for people; I admit I didn't think things through properly," Zap Actu GTA6 tells one critic. "Originally, I just wanted to create some kind of hype and see if people would catch on that it was AI. I wouldn't have thought it would go this far. Sorry to the people I disappointed." "For me it was more to amuse the community in reality, but oh well, it's not working," they say in another reply. "I'm not gonna force something that people don't appreciate." And yet... The account owner has now rebranded as Actuzz MAG, telling IGN they were "sorry for the false hope." Nonetheless, they have posted more AI-generated news speculating on "GTA 6: A Snow-Covered Vice City for Christmas 2026? A Winter DLC Fans Are Already Dreaming Of." Let's call it a lesson half-learned.

Amid Backlash, Creator of Viral GTA 6 Gameplay 'Leak' Video Insists It Was an 'Experiment' Designed to Show 'How Easy It Has Become to Blur the Line Between Reality and AI-Generated Content' - IGN

The person behind the viral GTA 6 gameplay "leak" video has admitted it was created using generative AI, amid a growing backlash from fans. IGN had reported on X / Twitter posts made by the Zap Actu GTA6 account, which included "leaked" gameplay clips of GTA 6 while pointing to a Discord. One video posted earlier this week -- now deleted -- showed playable character Lucia walking in the rain. It went viral, securing 8 million views in just over 24 hours despite a community note warning against trusting it as official footage from Rockstar. But there were many other similar clips, also with millions of impressions, from the same account, and based on the replies, a number of people believed they featured genuine leaked gameplay footage. In Zap Actu GTA6's Discord, a growing backlash emerged today as newcomers flooded in to seek clarity on whether these were genuine leaks of AI-generated videos. Now, ZapActu has come clean, issuing a statement and responding to questions from IGN. They insisted the videos were designed to "observe people's reactions and to demonstrate how easy it has become in 2025 to blur the line between reality and AI-generated content." ZapActu apologized "to anyone who felt frustrated, disappointed, or misled by these posts." They continued: "This was never done with bad intentions." ZapActu said they did not make any money from the posts, nor was there a financial motive behind "this experiment." ZapActu is now in the process of deleting posts and closing accounts. "My intention was never to harm anyone," they continued. "I simply wanted to create something intriguing that could bring people together and spark discussion within the GTA 6 community. I genuinely never expected a single video to generate such massive engagement and reach." And, in a direct message to IGN, they concluded: "It was a huge joke actually, I did it just to entertain the community. Sorry for the false hope lol." Misleading videos made by generative AI have exploded on the internet in recent years as the technology has become more popular and accessible. And it is a problem that affects all entertainment. Last month, IGN reported on physicist Brian Cox, who went public with complaints about YouTube accounts that had used AI to create deepfakes of him saying "nonsense" about comet 3I/ATLAS. Similarly, Keanu Reeves recently hit out at AI deepfakes of the John Wick star selling products without his permission, insisting "it's not a lot of fun." In July, it was reported that Reeves pays a company a few thousand dollars a month to get the likes of TikTok and Meta to take down imitators. Can anything meaningful be done? Last month, the Japanese government made a formal request asking OpenAI to refrain from copyright infringement after Sora 2 users generated videos featuring the likenesses of copyrighted characters from anime and video games. Sora 2, which OpenAI launched on October 1, is capable of generating 20-second long videos at 1080p resolution, complete with sound. Soon after its release, social media was flooded with videos generated by the app, many of which contained depictions of copyrighted characters including those from popular anime and game franchises such as One Piece, Demon Slayer, Pokémon, and Mario. Despite the protestations of the Japanese, OpenAI CEO Sam Altman has declared Sora 2 videos using copyrighted characters "interactive fan fiction." As for GTA 6, given the game isn't due out for another 12 months, expect more, increasingly convincing AI-generated gameplay "leaks" to hit the internet as fans desperate for official information from Rockstar - and, hopefully, Trailer 3 - wait on. If you're hunting for the best offers this week, we're actively rounding up the strongest Black Friday deals on video games, tech, and more. You can find all our top picks and price drops in our full Black Friday hub, or check out our relevant pages for PlayStation, Nintendo, and Xbox deals. Wesley is Director, News at IGN. Find him on Twitter at @wyp100. You can reach Wesley at [email protected] or confidentially at [email protected].

Increasingly Convincing AI-Generated GTA 6 Gameplay 'Leaks' Are Getting Millions of Views on Social Media - IGN

"GTA 6 ALERT - EXTREMELY SERIOUS SITUATION," reads the X / Twitter post from the Zap Actu GTA6 account. The supposed gameplay clip of Rockstar's upcoming surefire blockbuster is brief, but, if true, a shocking leak indeed. Of course, it's not true. It's not a leak. It's not even real gameplay. It's yet another AI-generated GTA 6 "leak" video viewed by millions that remains live across social media, which seems unable to do anything about it beyond the odd community note. The tweet below, published yesterday, November 25, has gone viral, securing 8 million views in just over 24 hours. Below it, a community note warns against its authenticity, but that appears to be doing little to dull its impact. And it is far from the only one. The same Twitter account responsible for this "leak" has published a number of similar clips in recent months, some of which which have a huge number of views, all in a desperate bid for followers and Discord members. Based on the replies, many Twitter users are taking these leaks at face value. It's a problem GTA 6 has faced for some time now, which probably comes as little surprise given the intense excitement and thirst for new information on what is expected to be the biggest entertainment launch of all time. But it is far from the only video game to suffer from this problem. Indeed, video games are not alone in this, either. Last month, IGN reported on physicist Brian Cox, who went public with complaints about YouTube accounts that had used AI to create deepfakes of him saying "nonsense" about comet 3I/ATLAS. Similarly, Keanu Reeves recently hit out at AI deepfakes of the John Wick star selling products without his permission, insisting "it's not a lot of fun." In July, it was reported that Reeves pays a company a few thousand dollars a month to get the likes of TikTok and Meta to take down imitators. In 2023, Tom Hanks warned fans that an AI version of his likeness was being used without his consent in an online advert for a dental plan. Last year, Morgan Freeman thanked fans who alerted him to AI-generated imitations of his voice online after a series of videos created by someone posing as his niece went viral. And in May this year, Jamie Lee Curtis was forced to appeal to Meta CEO Mark Zuckerberg in an Instagram post because she couldn't get the company to pull an AI-generated ad that featured her likeness for "some bullshit that I didn't authorize, agree to or endorse." What is the solution here? In July, YouTube was said to be preparing to update its policies to crack down on creators' ability to generate revenue from "inauthentic" content, which is made easy to produce on a massive scale with generative AI. While propelled forward by the great YouTube algorithm, you've probably stumbled across a fake trailer or two. The hope was that YouTube would be able to crack down on the channels that pump out this sort of low-effort content, but a cursory glance at YouTube shows this has yet to happen. Without legislation forcing content built by generative AI tools to include labels clearly marking it as such, or laws preventing deepfakes without permission, fans will continue to be misled by bad actors. And as generative AI technology improves, so it will become harder to distinguish between the fake and the real. Can anything meaningful be done? Last month, the Japanese government made a formal request asking OpenAI to refrain from copyright infringement after Sora 2 users generated videos featuring the likenesses of copyrighted characters from anime and video games. Sora 2, which OpenAI launched on October 1, is capable of generating 20-second long videos at 1080p resolution, complete with sound. Soon after its release, social media was flooded with videos generated by the app, many of which contained depictions of copyrighted characters including those from popular anime and game franchises such as One Piece, Demon Slayer, Pokémon, and Mario. Despite the prostetations of the Japanese, OpenAI CEO Sam Altman has declared Sora 2 videos using copyrighted characters "interactive fan fiction." Zap Actu GTA6, meanwhile, doesn't seem bothered by any potential backlash, and is posting even more fake GTA 6 "leaks." Their latest, published just a few hours ago at the time this article went live, already had 220,300 impressions. In their Discord, which is slowly gaining new members, one person who joined today questioned everything they were doing, saying: "If it's AI, that's disrespectful to the fans, friend @ZapActu." ZapActu has yet to respond. If you're hunting for the best offers this week, we're actively rounding up the strongest Black Friday deals on video games, tech, and more. You can find all our top picks and price drops in our full Black Friday hub, or check out our relevant pages for PlayStation, Nintendo, and Xbox deals.

AI-generated GTA VI gameplay clips are going viral, its creator calls it an experiment

GTA VI is one of the most awaited titles of the year and players across the globe are excited to find out whatever they can about Rockstar's next game. Perhaps that is why even the slightest hint of a GTA VI update, real or fake, instantly blows up online. Now, if you have been on X or Instagram recently, chances are you have already stumbled upon some GTA 6 leaked gameplay clips doing the rounds. The viral clips show rain-soaked streets, the new protagonist Lucia, a quick panshot of Vice City at night, and more. But guess what, all of these are completely fake and AI-generated. These videos look convincing enough to pass as leaked Rockstar footage and this kind of confusion is something that you would expect from the internet in 2025. The videos have left the GTA community split between amusement and frustration, but its creator says it was just an "experiment". Also read: Assassin's Creed IV: Black Flag remake could arrive soon: All you need to know A now-deleted post from the account Zap Actu GTA6 went viral and caused complete chaos on the internet. The post read "GTA 6 ALERT - EXTREMELY SERIOUS SITUATION," and featured what looked like a legitimate gameplay clip. In just over twenty-four hours, it racked up over eight million views. A community note was also added telling fans that the clip was fake, but the views kept climbing. At last, the people kept believing and the hype kept spinning out of control. This was not the first time the account had posted such clips. For months, it had been uploading short AI-made videos that blended into the feed just convincingly enough to fool half the internet. And it worked. People flooded the replies asking for more footage, more details and more "leaks," desperate for anything to ease the long wait for Rockstar's next trailer. As the backlash grew, the creator behind the videos finally came forward. In a lengthy post, Zap Actu GTA6 admitted everything was AI generated. There were no leaks or insider information in his videos. He called it just an attempt to "see how easy it has become in 2025 to fool people with fake leaks." The creator also insisted there was no money involved, so there were no sponsorships and no malicious intent. In the creator's words, it was "a huge joke" meant to entertain the community and spark discussion. In addition to this, it was also revealed that around half the viewers believed the footage was real, which they said was both surprising and worrying. The account is now deleting posts and winding down, but the fallout has opened up a much larger conversation about the online world we now live in.