Now That The Shoe Is On The Other Foot, Microsoft And Epic Games Suddenly Agree That AI Is Theft

The past four weeks have been something of a wake-up call for the video game industry. You’ve read the headlines, so you know the deal by now: AI advancements spearheaded by Claude Opus 5.5 have turned the world of coding on its axis. Decompilations that used to take months, or even years, are now being pumped out on a daily basis, and publishers can’t keep up with the deluge of vibe-coded browser ports and “native” PC recompilations. I’ve been covering and researching this over the past month, and there’s one recurring topic that keeps popping up in every comment section I read: karma. For many, this is the video game industry’s “you reap what you sow” moment, proof that the AI tools many publishers have brazenly admitted to using are a double-edged sword. It’s a tricky subject to tackle, because it comes packaged alongside so much bias. For instance, where the PlayStation emulation scene is concerned, many see it as retribution for Sony’s decision to kill physical media. Nintendo has maintained a mostly ambivalent stance on generative AI, yet some consider the explosion in vibe-coded recomps as punishment for the company’s litigious history. However, there’s a less ambiguous side to this discussion that I’m far more interested in tackling: hypocrisy. Many publishers are gleefully and very publicly advertising their use of AI to develop games, yet two specific publishers are attempting to have their cake and eat it too in this respect: Microsoft and Epic Games. Hundreds of DMCA notices have been fired off by companies in response to the recent flood of AI-powered recomps. Indie games are the prime targets, with takedown notices for open-sourced versions of Mega Crit’s Slay the Spire 2, Maddy Makes Games’ Celeste, and Ludeon Studios’ RimWorld all featuring in GitHub’s DMCA logs over the past four weeks. These three examples are important to note because they are blameless victims of vibe-coding: games that feature no generative AI content, yet have been open-sourced against their developers’ wishes. On the other hand, Epic Games issued a DMCA notice against a vibe-coded browser port of Fortnite on October 10, and Microsoft has issued multiple DMCA notices for Minecraft, Call of Duty, and Halo-related web browser ports over the past month. I could fill this article with dozens of examples of Microsoft’s and Epic Games’ love of AI, but here are two off the dome: Epic Games’ vibe-coded shitshow of a launcher that launched on October 14, and the mere existence of Microsoft’s Copilot. Epic Games took down a recent project that allowed Fortnite (recreation) to be playable through a web browser. They also DMCA (copyright) almost every post featuring it LMFAOO pic.twitter.com/7lSK4QksBk — Backers (@BackersGamesF) October 14, 2026 Here’s a fun quote from Tim Sweeney in an interview with PC Gamer: “It’s unfortunate that so many of the AI companies operating early on had such shitty practices, you know, like one of them was found by a court to have gone off to a BitTorrent site and downloaded terabytes of data. That’s ridiculous, they shouldn’t do that.” There’s an interesting use of the past tense here; AI companies had “such shitty practices,” not have. It’s almost as if he thinks this stolen data has been scrubbed and won’t recursively be used to continue training AI models or something. But AI remains built on theft, and no major AI companies have yet been forced to engage in model disgorgement, a legal remedy that forces AI companies to delete all models built on stolen data. Also note that Sweeney neglects to mention the exact AI company he’s referring to; he simply says that “one of them” was found to have used “a BitTorrent site” to train itself. As PC Gamer notes, Epic’s artists use both Google Gemini’s Nano Banana and OpenAI’s GPT Image. Both Google Gemini and OpenAI are involved in ongoing lawsuits that allege the scraping of materials from piracy sites. However, PC Gamer’s interview with Sweeney is from June 2026. Let’s give Epic the benefit of the doubt and assume they’ve moved on to an AI model that wasn’t built on stolen data. According to an October 12 report from Bloomberg, Epic Games’ upcoming “AI first” Disney/Marvel game will be developed using Anthropic’s Claude. In September 2025, Anthropic agreed to pay $1.5 billion for training its models using illegally obtained data sourced from BitTorrent. However, while Anthropic was forced to delete the stolen data, it was not forced to delete or rework the models it had trained on said data. Meanwhile, Epic Games’ Terms of Service specifically states that you “may not[…]use code or content extracted from the Licensed Products as training input for any Generative AI Program or as prompt-based input where the Generative AI Program trains on input data.” It seems Epic Games is happy to use AI models trained on stolen data, but it’s not happy for anyone else to train AI models on their stuff. And now, on to Microsoft—and honestly, where do you even begin here? Probably with the $80 billion per year the company has committed to building AI data centers. I guess that sum explains why Activision just had to use AI to generate in-game artwork for both Call of Duty: Black Ops 6 and Call of Duty: Black Ops 7. After all, if they didn’t save a few pennies by generating some AI slop, how would Microsoft’s shareholders know they’re getting their money’s worth? Speaking of Call of Duty, did you notice how many vibe-coded CoD web ports were wiped from the internet via DMCA notices over the past month? I count at least six. To be fair, Xbox CEO Asha Sharma stated the company “remain[s] committed” to “being responsible stewards of IP” in her Xbox Labs announcement on October 13, so I guess we shouldn’t be surprised. Today we're standing up XBOX LABS (https://t.co/wzXPNnCmI6). Here is the note that I sent to the team: Team,The first video games emerged some 75 years ago. In the time since, there have been booms and busts, but every year there are more great games to play, more devices to… — ASHA (@asha_shar) October 13, 2026 But let’s be honest for a second: “being responsible stewards of IP” isn’t Microsoft’s modus operandi. “Responsible” is just a cute buzzword to throw in here, as it implies that the company gives a shit about everyone else’s IP, too. Funny, considering that time Microsoft’s Senior Product Manager suggested users use a “well-known dataset”—all seven Harry Potter books—to “generate new AI-driven Harry Potter fan fiction.” That’s just a one-off example, though. It’s not like Brent Hecht, Microsoft’s director of applied sciences, stated that the scraping the internet by LLMs represents the “largest theft of labor in human history” or anything, right? In case you didn’t know, The New York Times is currently suing both Microsoft and OpenAI, and Hecht’s comments were obtained during discovery. “Millions of people around the world will soon consider large models ‘hoovering up’ all their work to be an astonishing theft of unprecedented proportions,” stated Hecht. “Almost no one intended for content they created to be used in this fashion, nor are they compensated for its use.” There’s a depressingly humorous quote in there too, sourced directly from Microsoft’s internal communications: a warning that generative AI could “significantly disrupt the employment of the very people who generated the data on which the foundation model was trained.” “Our AI content strategy has started a ‘doom loop’ that will hurt the performance of our models and the entire web at the same time,” stated another internal Microsoft memo. “It is highly unusual that an end-product threatens the economic foundations of its essential suppliers, but that is the situation we have created for our LLM business with respect to its ‘content supply chain.’” Did you also know that Microsoft Chairman and CEO Satya Nadella, along with Vice Chair and President Brad Smith, have suddenly become very concerned with tighter AI regulations over the past month? Brad thinks it’s worrying that “your washing machine is subject to more safety rules than AI,” and Nadella now believes that an “authorized person” should be able “to pause or shut down a model mid-task.” Are these quotes related to web browser ports of Microsoft IPs? Not specifically, but it’s awfully funny that they’re changing their tune now that AI is empowering individuals to commit IP theft. Guess it’s fine when they do it, though! Let me be clear: I don’t think it should be legal for you to vibe-code ports of Fortnite and Call of Duty, and I’m certainly not against better AI regulations. As much as I’m enjoying seeing them get a taste of their own medicine, I’m not advocating for you to rip off the big corps. I’m not asking you to “leave the billion-dollar company alone”—to be frank, I couldn’t give less of a fuck if Microsoft and Epic Games’ profit margins dip. But AI’s ability to recompile games isn’t hurting the big guys; it’s hurting indie games. AAA publishers can just throw lawyers at this over and over again, but for indie developers who self-publish their own games, such efforts would burn a sizable hole in the company account. I would be a hypocrite if I advocated for vibe-coded theft, which hurts indie developers more than it does billion-dollar companies. This article isn’t in defense of Microsoft and Epic Games; it’s in support of Mega Crit, Maddy Makes Games, Ludeon Studios, and every indie dev out there that’s suddenly too afraid to publish their demos on Steam. You didn’t invest in the stealing machine. You didn’t use AI to cut corners. You didn’t agree to this. You have my sympathy.