Vietnam’s AI Copyright Clarity: Who Owns the Art When the Artist Is a Machine?

Imagine spending months training an AI model, only to discover that the masterpiece it generates belongs to… no one. Or worse, that you’re liable for copyright infringement because your AI scraped a protected dataset. That nightmare just got a bit more real—and a bit more clear—thanks to Vietnam’s new guidelines on AI and copyright.

What Vietnam Just Did

Vietnam’s Ministry of Culture, Sports and Tourism released official guidance clarifying three thorny issues: who owns AI-generated works, what training data is fair game, and who gets sued when an AI copies something it shouldn’t. The rules are practical, not philosophical—they aim to give businesses a roadmap, not a debate on machine consciousness.

Authorship: The Human Touch Still Matters

Under the new guidelines, only works created with “significant human creative input” qualify for copyright protection. If you just type a prompt and let the AI do the rest, you’re out of luck. But if you curate, edit, and shape the output—like a photographer directing a shoot—you may own the rights. It’s a bit like claiming you baked a cake because you pressed “start” on the mixer. Sorry, the mixer doesn’t get a slice.

Training Data: The Wild West Gets a Sheriff

Vietnam now requires that training datasets be sourced from lawful origins. Using copyrighted material without permission? That’s a risk. The guidelines suggest using public domain, licensed, or synthetic data. Think of it as grocery shopping for your AI: you wouldn’t steal ingredients from a neighbor’s garden, so don’t scrape protected content without a license.

Liability: Who Pays When the AI Messes Up?

If your AI generates infringing content, you—the developer or deployer—are on the hook. Vietnam follows a strict liability model: ignorance of your training data’s provenance is no defense. This is a wake-up call for companies that treat AI as a black box. You need to audit your data, document your sources, and have a takedown process ready. Because when the AI copies a Disney character, Mickey’s lawyers won’t care that you didn’t know.

A Global Perspective: How Vietnam Compares

Vietnam’s approach aligns with the EU’s Copyright Directive, which also requires human authorship and allows text and data mining for research but not commercial use. The US Copyright Office has similarly denied registration for purely AI-generated works. China, on the other hand, has granted copyright to AI-assisted works with significant human input. Vietnam sits comfortably in the middle—pragmatic, not pioneering.

What This Means for Your Business

If you develop or deploy AI in Vietnam—or anywhere with similar rules—start now. Audit your training data. Document human involvement in AI outputs. Update your terms of service. And for heaven’s sake, don’t assume your AI is a free pass to copy the internet. Reading these guidelines is less painful than cleaning grout with a toothbrush, I promise.

The bottom line? Vietnam just gave the AI industry a gift: clarity. Use it wisely, or prepare for a world where your AI’s creativity becomes your liability.