For the longest time, lawyers have been the gatekeepers of legal knowledge.

We are the ones people turned to if they wanted to understand the law, interpret a judgment or draft an agreement. That knowledge was held first in private libraries guarded by dusty hardbacks, then in specialist databases that you needed a password (and a law degree) to navigate.

Then the internet came along and cracked open the vault. Now? You can Google the law. You can read judgments online. You can download a template agreement in seconds.

And yet, something even more transformative is now reshaping everything: generative artificial intelligence or Gen AI. It’s not just another research tool or fancy upgrade to the legal database. Gen AI can now summarise law, draft arguments, interpret regulations and structure advice. The key reason it’s able to do this so well is because of how it works: at its core, AI predicts the next most likely word. That might sound simplistic, but that’s exactly what makes it powerful.

Legal text, for the most part, is incredibly predictable. It’s structured, formal and non-creative. We spend hours perfecting a lease or a shareholder agreement and the truth is, majority of the language is boilerplate. It’s the kind of stuff AI loves!

Some Serious Questions

So, what happens when this kind of intelligence becomes widely accessible? When the flood gates are opened and our legal knowledge is not considered so "valuable"? How do we justify our fees?

The uncomfortable truth is, many won’t be able to. If your day job revolves around drafting straightforward documents, reviewing basic contracts or writing generic advices based on publicly available law, you’re going to find it increasingly hard to compete with machines that can do the same thing in seconds. This is particularly relevant for graduate or junior lawyers. Many AI tools I use can already draft at or above a junior lawyer level.

It’s not that your work isn’t valuable. It’s that the market has moved and the perceived value has changed.

And the numbers back that up. A 2023 report by Goldman Sachs estimated that 44% of legal jobs are at risk of being automated. That’s nearly half the profession. When you break that down, it’s easy to see where the axe might fall first. Everyday tasks—lease reviews, contract templates, basic advice emails, procedural letters—all these things are easily replicated by tools like ChatGPT and more advanced legal-specific Gen AIs.

And clients are beginning to realise this. They’re experimenting. They’re learning. Some of them are even comparing your advice with the AI’s just to see how close it is.

Are Lawyers Done?

This all reminds me of something very real from my own life. Before becoming a lawyer, I was a pharmacist. As a pharmacy student and later a graduate, discount pharmacies started popping up everywhere. We were told that patients would still value service, advice and the relationship with their pharmacist.

That assumption didn’t hold. Most people wanted cheaper meds. That was it.

And I choose to apply that same reasoning to all the articles currently being written on the subject. The mostly upbeat pieces that only talk about how Gen AI will add to the services being provided by lawyers. The fact that clients will only want human interaction and relationship building. I don't believe that.

In hindsight, the patients who came to us and paid more were paying for the "human" experience—the conversation, the reassurance, the sense that someone actually cared. But that wasn’t enough to stop the rise of the discount model. The majority of pharmacies eventually adapted or disappeared.

And here’s where the comparison gets a bit grim for the legal profession. People will only compromise so much on their health. But legal rights? Most clients don’t even realise what they’re giving up when they choose not to engage a lawyer. They don’t know what to value, or why it matters, until something goes wrong. That’s why I think lawyers are even more vulnerable to this shift.

Time to Adapt

But this isn’t meant to be a doomsday piece. Because while Gen AI is very good, it isn’t everything. There are still things it can’t replicate, at least not yet. It can’t read a room. It can’t understand a client’s emotional state. It can’t navigate grey areas where the law is uncertain and strategic judgment is required. It can’t handle multi-disciplinary problems where the answer requires not just knowing the law, but knowing how to apply it commercially, tactically or politically.

That’s the kind of lawyer that will thrive. The ones who combine specialist knowledge with empathy, who can operate across boundaries like merging commercial law with regulatory strategy or IP law with digital innovation. These are the lawyers who won’t be replaced so easily, because their value doesn’t come from what they know but how they think and who they are.

So, I think we will start seeing the slow decline of specialist lawyers and the rise of more generalists. Law firms that focus on narrow areas of law—conveyancing, leasing, strata, immigration—will find it more difficult to adapt to the change.

What Now?

And there’s something else. AI isn’t going away. It’s becoming a tool we all need to learn. In the same way we learnt how to use legal databases in law school, we now need to learn how to work with AI. That means learning how to prompt it properly. Understanding its limitations. Cross-checking and verifying its outputs.

I’ve spent the last 12 months building that muscle. Playing with different tools, trying different inputs, refining prompts, seeing how it interprets nuance. And I can tell you that I agree there will be a huge gap between lawyers who embrace AI and those who don’t.

When thinking about this topic, I remembered something from my first few weeks of law school. We were in the university computer lab, being shown how to use legal research platforms. One of the tutors said something along the lines of: "You’re not here to memorise the law. You’re here to learn where and how to find the information that helps your client."

That was over a decade ago but now seems more relevant.

At the end of the day, AI isn’t the enemy. But it does force us to confront something I believe the legal industry has ignored for too long: legal knowledge alone isn’t enough. What matters now is how we use that knowledge, how we connect with people and how we adapt to the tools shaping our world.

Lawyers will need to become more "human" than ever before.