AI Won’t Replace Junior Lawyers. It Will Give Them Better Work

At every law firm leadership conference I attended this year, the same prediction kept coming up: Artificial intelligence will automate the work partners traditionally assign to junior associates. Therefore, the theory goes, the traditional law firm pyramid will flatten, and we will need far fewer junior lawyers.

I disagree.

AI will create more work for firms, including our most junior lawyers, and let us deliver the services clients want at lower cost and higher quality. I have never met a client who came to us looking to buy billable hours; they want solutions to their business problems, and the billable hour has long been a proxy for how we charge for those services.

In some cases, the hourly rate model has priced sophisticated legal work out of existence because clients could not justify the cost. AI changes that, and I believe it changes it in our clients’ favor.

In the 1860s, the economist William Stanley Jevons observed that as steam engines grew more efficient and coal grew cheaper to use, demand for coal didn’t fall. Instead, demand for coal exploded, since cheap coal was now worth using in far more places. The same logic, known as the Jevons Paradox, should apply to legal services.

As work becomes faster and cheaper, clients will buy more of it, not less. In practice, that means matters to clients once deferred, narrowed, or abandoned can suddenly become economically viable and move from a client’s “nice to have” bucket to the active docket.

I believe the surge in work will come from two directions. First, the cost of the routine parts of legal work is falling, sometimes dramatically, even as the value of the judgment around that work keeps rising. We’re already seeing this in parts of our firm.

In structured finance, we use large language models to review high volumes of trust and servicing agreements, surfacing our clients’ obligations and strengthening our hand in negotiating deals. In our funds practice, we have largely automated side-letter review across funds with hundreds of investors. And in litigation and disputes, our lawyers use AI to comb through hundreds of deposition transcripts for the quote that supports, or undermines, an argument in a brief.

A few years ago, a client looked hard at a large project, saw the price tag, and decided not to proceed. This past year, with AI assistance we fully disclosed, we brought the expected cost of that same project down by roughly 70%, even though our hourly rates had risen in the interim. After AI, we had real revenue at higher rates and a client that had limited that risk at a far lower cost.

The same logic, applied to discovery, can take a middle-market dispute for which document review once cost millions and dramatically bring down that cost, making an entire tier of meritorious litigation once again viable. As these tools mature, we will share those gains with our clients not just through lower effective hours but through the faster implementation of alternative fee arrangements that tie total cost more closely to value.

The second factor that I believe will lead to increased demand for law firm services is that our clients will become more productive. For example, the private fund managers we represent will use AI to vet more deal flow and negotiate more private equity and venture transactions each year, and more transactions mean more legal work.

Our technology and life sciences clients will generate ideas, and intellectual property worth protecting, faster than ever. And AI itself will generate a steady stream of disputes for our litigators and ultimately create significant work for our bankruptcy and restructuring teams.

Some of these extra demands will be offset. Our clients will have access to the same legally trained AI tools we rely on, and in-house counsel will use them to pull more routine work back inside the business.

Most in-house work will be high‑volume, low‑judgment tasks that was already turning into commodities. The same productivity gains that let a client absorb routine tasks also allows that client to chase more deals, protect more IP, and end up in more disputes. The higher-stakes matters that follow are exactly the ones that call for seasoned judgment, not just a capable AI agent. Knowing current market terms in private deals or which provision matters isn’t something a client can download.

Where do junior associates fit in all of this? I asked a number of our counsel and associates how our youngest lawyers are using these tools.

On smaller matters, senior lawyers can sometimes just do the junior-level task themselves, but on the larger and more complex matters, AI makes our junior lawyers more productive while giving them something I never had, a co-pilot they can engage with to explain an unfamiliar provision or a theory of a case that will improve their training and their ultimate work product. But, if demand for legal work rises faster than AI alone can absorb it, and our juniors can take on work that once demanded a more senior lawyer, a firm doesn’t need fewer junior lawyers to meet that demand. It needs more of them, trained to expertly use and learn from AI tools that will allow them to take on higher-value work sooner.

This all only works if two things happen. First, junior lawyers must keep thinking for themselves and questioning the output. Second, supervising lawyers must do their part by explaining how a matter is structured, sharing prompts and best use cases, and treating the technology as something to teach rather than hoard.

I get excited about connecting junior associates to our firm’s “golden source” of precedent, a curated store of our best documents with tools that highlight typical deal terms and litigation strategies. An associate with that access can tap the collective knowledge of the whole firm in a way no client can replicate. That proprietary know‑how will matter most as the underlying AI models themselves become commodities.

In the next few years, law school graduates will be increasingly native in this technology. Older lawyers have always learned new tools from younger colleagues, and many of today’s junior attorneys arrive with real work experience and sharper judgment than earlier generations.

The conventional wisdom has it backwards. AI doesn’t so much replace the junior lawyer as raising the floor on what a junior lawyer can do, and the client is the one who benefits.

This article does not necessarily reflect the opinion of Bloomberg Industry Group Inc., the publisher of Bloomberg Law, Bloomberg Tax, and Bloomberg Government, or its owners.

Author Information

Gary Wingens is chair of Lowenstein Sandler. He has two sons in legal practice and a third entering law school this summer, which gives him a personal stake in the optimistic view expressed in this piece.

Interested in writing? Review our author guidelines, and submit pitches to Insights@bloombergindustry.com.