You’ve probably noticed the tech industry has a massive blind spot when it comes to intellectual property. If you download a movie from Netflix and upload it to The Pirate Bay for people to watch for free, you are a criminal. You get sued, fined, and maybe even jailed. But if a multi-billion dollar AI company scrapes millions of books, articles, and movies to train a model—and then charges users a monthly subscription to access that model—that’s called “innovation.”
The exact same act of unauthorized copying is a felony when done for free, but a business model when done for profit.
It’s a blatant paradox. On one hand, we have draconian piracy laws designed to protect corporate copyright holders. On the other, we have AI behemoths like Anthropic and OpenAI vacuuming up the internet’s collective creativity, claiming it’s just “fair use.” How did we end up in a reality where the economic outcome—profit versus free access—dictates what is legal?
The truth is, this isn’t about ethics or technical nuances. It’s about power. When Napster threatened the music industry, lobbyists crushed it. When Aaron Swartz tried to make academic papers free, the legal system threw the book at him. But when Silicon Valley’s newest darlings needed data to build their trillion-dollar machines, the rules magically bent.
Piracy laws were written before machine learning existed, leaving a regulatory gap that conveniently favors deep-pocketed corporations.
AI companies successfully framed model training as “transformative use.” They argue that because the AI doesn’t spit out a direct, identical copy of your novel or screenplay, no harm is done. But let’s be real: if I steal your car, strip it for parts, and build a motorcycle I sell for a profit, I didn’t “transform” your car. I stole it and monetized it. The fact that the output looks different doesn’t erase the theft of the input.
If you create anything—if you write, code, paint, or film—you should be furious. The system is rigged to protect the platforms distributing the work, not the people making it.
We have built an economy where stealing from the rich gets you sued, but stealing from millions of individual creators gets you a Series C funding round.
The next time an AI executive talks about democratizing knowledge or pushing the boundaries of human potential, remember the double standard. They aren’t democratizing anything; they are privatizing the commons. They took your work, put it behind a paywall, and convinced the courts it was an accident of progress.
FAQ
Q: Isn't AI training technically 'transformative use' under copyright law?
A: Only because deep-pocketed legal teams successfully lobbied to frame it that way. If you strip a car for parts to build a motorcycle you sell, it's still theft, even if the final product looks different.
Q: What does this mean for everyday creators?
A: Your work is being monetized without your consent or compensation. The platforms hold all the cards, and the legal system is currently designed to protect them, not you.
Q: So, should we just shut down all AI development?
A: No, but we need to stop pretending this is a legal gray area driven by innovation. It's a power grab. We need licensing models that actually pay creators, not just legal loopholes that protect tech monopolies.