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

Hypocrisy has a way of showing up right when it becomes inconvenient, and the latest turn in the debate over generative AI is a textbook example. Two of the biggest names in tech and gaming, Microsoft and Epic Games, have spent years embracing, funding, and promoting artificial intelligence. Now that the technology is being used in ways that affect their own products, they have suddenly found strong feelings about copying, scraping, and theft. Rules for thee, but not for me.

I should say up front that I'm working from the headline and the framing, not from the full text of the original Kotaku report. So I'll focus on the broader pattern and the arguments it raises, and I won't invent specific quotes or details I can't verify. Treat what follows as commentary on that pattern.

How the Shoe Ended Up on the Other Foot

For the last few years, creators of every kind have raised the alarm about AI models trained on their work without permission. Artists, writers, voice actors, and musicians have all argued that feeding their output into a model and then selling the results is a form of appropriation. The response from much of the tech industry has been remarkably consistent: it's fair use, it's transformative, and it's the future.

That stance was easy to hold while the people complaining were individual creators with limited legal resources. It gets harder when the complaint comes from inside the house. When a company's own assets, tools, or content are the ones being copied by AI, the tone changes quickly.

That is the heart of the story. The principle of "you shouldn't take what isn't yours" is not new, but it seems to become much more persuasive once a large corporation is the one being taken from.

Microsoft's Complicated Relationship With AI

Microsoft is one of the most aggressive backers of generative AI in the world. Its investment in OpenAI, its Copilot branding across seemingly every product, and its push to put AI into Windows, Office, and GitHub all signal a company that believes in this technology wholeheartedly.

GitHub Copilot in particular drew criticism from developers who said it was trained on open-source code without honoring licensing terms. Microsoft defended the approach, and the matter ended up in court. The company has long argued that learning from public material is not the same as stealing it.

So when Microsoft raises concerns about AI misuse that affects its own interests, it's fair to ask whether those concerns apply equally to everyone else. If copying is theft in one case, it's hard to argue it's innovation in another.

Epic Games and the Ownership Question

Epic Games has a different but equally interesting history. As the maker of Fortnite and Unreal Engine, it has positioned itself as a champion of creators and an opponent of what it sees as unfair gatekeeping by platform holders. Its public fights with Apple and Google made Tim Sweeney a vocal advocate for open markets and developer rights.

At the same time, Epic has embraced AI features within its own ecosystem, including tools that let creators and even in-game characters use generative technology. That puts the company in an awkward position when the conversation turns to who gets to use what, and under which conditions.

Defending creators is a strong brand. But it requires consistency, and critics are quick to point out when a company's principles seem to shift depending on who benefits.

The Double Standard Critics Keep Pointing Out

The charge of a double standard is straightforward. If scraping creative work to train a model is acceptable when a large company does it, then it's difficult to condemn others for doing the same to that company. If it is theft when it happens to Microsoft or Epic, then the same logic applies to the illustrators and writers who spoke up first.

Many observers see this as proof that the real dividing line isn't ethics but power. Those with legal teams and market leverage get to define what counts as theft, while everyone else is told to adapt.

That doesn't mean the companies' complaints are necessarily wrong. It means their credibility is damaged by years of arguing the opposite when the shoe was on the other foot.

What This Means for Artists and Developers

For working artists and developers, this moment is equal parts vindication and frustration. It's validating to see powerful companies acknowledge that unauthorized copying causes harm. It's frustrating that it took their own interests being threatened to get there.

Many creators have said for years that they weren't against technology itself. They were against being used as raw material without consent, credit, or compensation. That is a narrower and more reasonable complaint than the industry often suggests.

If these companies now accept that principle, there's an opportunity to push for real standards, such as opt-in training data, licensing deals, and transparency about what goes into these models.

The Legal Gray Area Still Looms Large

The law has not caught up with any of this. Courts around the world are still working through whether training on copyrighted material counts as fair use, and early rulings have pointed in different directions. Until there is clearer precedent, every company is effectively making its own bet.

That uncertainty explains some of the inconsistency. Companies argue whichever position protects them in a given dispute, hoping the final outcome favors their long-term strategy. It's a pragmatic approach, but it is not a principled one.

Anyone hoping for a tidy resolution should prepare to wait. These cases could take years, and the outcomes will shape how creative work is valued for decades.

Why Consistency Matters in the AI Debate

Consistency isn't just a talking point. It determines whether the public can trust companies to act in good faith as AI becomes more embedded in daily life. When a corporation changes its argument based on convenience, it signals that its stated values are negotiable.

The gaming industry is an especially important arena for this debate. Games are built on the labor of huge teams of artists, writers, designers, and performers, many of whom already worry about job security. Treating their work as free training data adds insult to that anxiety.

If big players want to be taken seriously when they say AI can be theft, they will need to apply that standard to their own practices, not only to those who copy them.

Final Thoughts on Rules for Thee, Not for Me

The story here is less about any one company and more about a pattern. When copying benefits you, it's progress. When it costs you, it's a crime. That kind of reasoning rarely survives scrutiny, and the public is increasingly attuned to it.

Still, there's a hopeful reading. If even the strongest AI boosters now agree that taking without permission is a problem, then the foundation exists for fairer rules. The question is whether they'll extend that fairness to everyone, or only to themselves.

Until then, expect more of the same: loud principles, flexible application, and a lot of people pointing out that the shoe looks very different depending on which foot it's on.

If you have any questions, please don't hesitate to Contact Us