How A Computing Academic Actually Beat His University Using Ai

How A Computing Academic Actually Beat His University Using Ai

Gregory Baker just changed the way we think about legal representation. He didn't hire a high-priced firm. He didn't spend months in a conference room with senior partners. Instead, he built his own digital legal team using AI.

As a computing academic at Macquarie University, Baker was frustrated. He felt he deserved a permanent part-time role rather than his casual contract. The university disagreed. So, he took them to the Fair Work Commission. He won.

Most headlines claim an "AI legal team" won the case. That’s a massive oversimplification. It ignores the reality of how these tools actually function. Baker didn't just ask a chatbot to win for him. He treated those agents like interns. He guided them. He checked their work. He applied his own domain expertise to their output.

The truth about AI legal success

If you think you can walk into a courtroom tomorrow with a generic prompt and walk out with a win, you’re mistaken. Most people who try to use AI for legal tasks end up drowning in "hallucinations." They present nonsense. They cite fake cases. Judges hate it.

Baker succeeded because he understood two things the average person ignores. First, he limited his scope. He didn't try to solve the entire employment law code. He focused on a specific, narrow provision regarding casual conversion. Second, he managed his AI agents with the same rigor he uses for software development.

He didn't just use one tool. He used multiple paid agents, like ChatGPT Pro, to cross-reference his arguments. He had them look for holes in his case before his employer could find them. Think of it as a low-cost, high-intensity red teaming exercise.

Access to justice is the real story

The most important takeaway isn't that robots are replacing lawyers. It’s that technology is finally lowering the barrier for the "missing middle."

Millions of people fall into a frustrating gap. They make too much money to qualify for free legal aid, but they can't afford to pay a top-tier firm to fight a relatively straightforward dispute. This is where AI actually shines. It’s perfect for simple, rules-based arguments where the facts are well-documented.

The Fair Work Commission is designed for self-representation, but navigating it still feels daunting for most. Baker proved that if you are organized and have the right tools, you don't need a JD to make a compelling argument. He basically automated the grunt work of document preparation and reference checking.

💡 You might also like: laptop screen protector for eyes

Don't try this without a plan

If you are facing a similar dispute, don't rush into it thinking the software will do the heavy lifting. You will fail. If you want to use these tools effectively, you need to follow a different path:

  • Narrow your focus. If your claim is messy or covers too many variables, the AI will get lost. Stick to one clear, provable argument.
  • Treat the AI like a junior researcher, not a senior partner. Never take the first answer. Always ask the AI to "check its work" against specific documents you provide.
  • Verify every single claim. If the tool cites a statute, go find the official government website and read the text yourself. Do not trust the AI's summary.
  • Use local context. Make sure you feed the AI your specific jurisdiction’s rules. A general model trained on US law will provide useless advice for an Australian tribunal.

The changing future of tribunals

Courts and tribunals are already feeling the heat. They are dealing with a massive surge in case numbers. Some sources say workloads for bodies like the Fair Work Commission have jumped 70% in three years.

Some of this is due to people being more aware of their rights. But a lot of it is also just easier access to filing systems. When you make it easier for people to bring a claim, you get more claims. Not all of them are valid. Some are absolute messes.

We’re going to see a clash between the democratization of legal arguments and the administrative burden on our judicial systems. Don't be the person who adds to the pile of garbage. Be the person who uses these tools to present a crisp, precise, and well-researched case.

Stop looking for a "magic button" that wins cases. Start learning how to structure your arguments so clearly that even a machine can help you build them. If you’re in a dispute, get your documents organized, define your narrow scope, and start testing your arguments against the AI before you ever step foot in a hearing room. That’s how you win.

JW

Jun Wood

Jun Wood is a meticulous researcher and eloquent writer, recognized for delivering accurate, insightful content that keeps readers coming back.