This is quite interesting.

Today, I came across a Bloomberg Law report suggesting that the copyright issues surrounding AI-generated code (they call it “Vibe Coding”) might soon blow up. In simple terms, when you use natural language to instruct AI to write code, who owns the final output? It sounds technical, but it’s actually a legal minefield.

Honestly, this debate is long overdue. Who hasn’t used Copilot or ChatGPT to write a few lines of code these days? But have you considered that AI-generated code might not be “original” at all—it could be cobbled together from open-source projects or even directly copied from copyrighted snippets. Developers often have no idea, since the AI won’t volunteer, “Hey, I lifted this from the third answer on Stack Overflow.”

The legal side is even more surreal. Traditional copyright law requires code to be written by humans to count. Now, with AI acting as a “middleman,” copyright offices are stumped—last year, the U.S. Copyright Office rejected an application for AI-generated artwork with the blunt reasoning: “No human author.” But is code the same? If you ask AI to write a sorting algorithm and it ends up identical to a textbook example, is that infringement or just “human consensus”?

The open-source community is probably sweating too. They’ve mastered licenses like MIT and GPL, but now there’s an AI wildcard. For example, if you use AI to modify a GPL project, does the license require you to open-source it? But the AI never “agreed” to the license! If this ever goes to court, the legal fees alone could fund a unicorn startup.

Technically, it’s darkly comedic. Today’s large models are trained on datasets that likely contain some “dirty” code, but no one knows how much. It’s like eating a burger without knowing which animals are in the patty—but if you accidentally consume a protected species, who’s liable? If developers get sued for copyright infringement over AI-generated code, is that fair?

The wildest part? Policies are still playing catch-up. The EU AI Act and the U.S. Copyright Office are in wait-and-see mode, with barely any legal precedents. The result? Companies are racing to boost efficiency with AI while their legal teams pray silently. One CTO I know told me they now manually annotate every AI-generated code block with: “May contain AI ingredients, use at your own risk”—like a cigarette warning label.

That said, this bomb will detonate eventually. Either a massive lawsuit will shock everyone awake, or the open-source community will invent an “AI Code Ethics License.” But my bet? People will pretend nothing’s wrong until some unlucky soul gets hit with a class-action suit.

(Original article link here: Bloomberg Law report [hypothetical])

What do you think? Does your company have risk controls for AI-generated code? Drop a comment and let’s chat.