How Australia's Kate Chaney Wants to Make AI Training Simpler (and Why It's Complicated)

Independent MP Kate Chaney has proposed a way to make it easier for AI companies to get permission to use published content for training their systems. Instead of negotiating with thousands of individual writers, publishers, and artists, she suggests AI companies should be able to make a handful of deals with larger groups that represent many creators at once. Australia is currently trying to figure out how to regulate AI while also keeping room for innovation — and Chaney's proposal sits right in the middle of that debate.

The Problem: Millions of Negotiations, One AI Model

Here's the core issue Chaney is trying to solve. To train modern AI systems, companies need vast amounts of text, images, and other content. But that content usually belongs to someone — a novelist, a newspaper, a photographer. Right now, AI companies are supposed to ask permission from each copyright holder individually. That means potentially thousands of separate negotiations, each one requiring lawyers and time.

Companies with big legal budgets can handle this. Smaller AI startups cannot. This creates a bottleneck that can slow down development and end up concentrating power in the hands of a few wealthy firms.

Chaney's solution borrows from how the music and broadcasting industries already handle this problem. When a radio station wants to play music, it doesn't negotiate with every songwriter and record label separately. Instead, it buys a blanket license from organizations like performance rights societies, which negotiate on behalf of thousands of artists. Chaney is suggesting something similar could work for AI: AI companies would negotiate with publishers' associations, creative industry groups, or government licensing bodies that represent many content creators.

What's Happening in Australia Right Now

The timing of Chaney's proposal matters. In 2024, the Australian Government signaled it might be willing to rethink copyright laws as part of discussions with AI companies — though the details were kept quiet. More recently, Parliament held hearings in February 2026 about labeling content created by AI, which suggests lawmakers are thinking about both sides of this problem: making AI development easier while also making sure people can tell which content is human-made and which is AI-generated.

This two-pronged approach reflects a difficult balancing act. Countries around the world are wrestling with the same question: how do you encourage innovation while protecting the people whose work feeds that innovation?

Australia has some weight in these global conversations. As a significant middle-power economy, how it handles AI regulation can influence what other countries do. If Chaney's model works, other countries might copy it.

Chaney's Bigger Picture on AI

The content licensing idea is just one part of how Chaney thinks about technology regulation. She has pushed for stronger AI safety rules in Australia while also focusing on related issues like protecting people from online harms and regulating online gambling. Instead of choosing between "let innovation happen" or "clamp down on tech," she's trying to address multiple concerns at the same time.

Her track record shows this pattern. She has proposed changes to gambling laws and participated in parliamentary inquiries on technology and elections — work that suggests she sees tech policy as interconnected. You can't just handle one problem in isolation.

What Could Actually Go Wrong (and Right)

If this licensing system gets built, there are potential benefits. AI companies could spend less time on legal negotiations and launch products faster. For smaller startups, lower compliance costs might make it easier to compete.

But there are real questions that need answering. Who decides what content creators get paid and how much? What can companies actually use the data for — just training models, or selling them too? How is this system enforced? And here's a trickier one: even if licensing becomes simpler, will it actually help smaller companies, or will it just make things easier for the big players who already have relationships with major publishers?

The European Union has its own AI rules, but most content licensing is still handled by individual countries' copyright laws. The United States mostly relies on a legal concept called "fair use" — which is basically a gray area where companies and creators end up fighting in court. This lack of clarity means it's hard to know what approach would work best.

Why This Matters Beyond Australia

Technologies often force societies to rethink how ownership works. When photocopiers became cheap and widespread, people had similar debates about copying books. The same thing happened with cassette recorders and file-sharing websites. Each time, the solution wasn't a total victory for either "creators' rights" or "easy access" — it involved both sides compromising.

The broader context here suggests Australia may be trying to position itself as a middle option. Not as restrictive as Europe or as hands-off as the United States. If that's successful, it could attract AI companies to Australia while also giving the country credibility as a thoughtful voice in global AI governance discussions.

The real test, though, won't be the idea itself — it will be the details. How are creators compensated? What fees do AI companies pay? Who oversees the system? These practical questions will determine whether streamlined licensing actually works fairly, or just shuffles the same power imbalances around in a different form.

Ultimately, this matters beyond Australia's borders. A handful of companies dominate AI development globally, and they're always thinking about where to set up operations and where to get their training data. National licensing frameworks like Chaney's proposal could actually change those calculations. It's one possible model among several, and its success depends on whether lawmakers follow through and whether the AI industry is willing to adopt it.