AI reshaping productivity inside Australian law firms

The shift is already underway, with top-tier law firms like Glibert +Tobin, Holding Redlich, Clayton Utz and Allens just some of the big names embedding the latest AI in their firms.

Additionally, there are early signals GenAI is impacting productivity in the sector by sparking a redistribution of work inside law firms. That is according to Thomson Reuters in its latest Australian legal market update, which says productivity remains stable at local law firms, but notes non‑equity partners and associates are logging fewer hours, while senior associates and equity partners are working more.

Non-equity partners recorded a third straight productivity decline in the first half of 2025/26, it says, with hours worked by junior and mid-level associates also slightly down. This was offset by senior associates and equity partners who logged more hours, keeping total numbers stable across the country, as per the update.

A possible explanation for the trend, it says, is GenAI now widely being used on research, drafting, and document review – tasks historically done by junior and mid-level associates. “This is precisely the pattern we would expect,” it says. “The distribution of hours may represent an early sign of how AI is beginning to reshape the traditional leverage model.”

In Sydney, Zed Law principal Nandan Subramaniam says the research is on point, suggesting that the results capture the “early structural impact of generative AI” in the field.

“The types of AI driving this are tools built on LLMs (large language models) – like GPT-4, Claude, and purpose-built legal AI platforms – which are now being deployed for legal research, first-pass document review, contract drafting and mark-up, due diligence, and knowledge management,” Subramaniam tells LSJ Online.

“These are precisely the tasks that have historically filled the hours of junior lawyers and non-equity partners.

“As those tasks are completed faster or partially automated, those fee earners naturally log fewer hours on a per-matter basis.”

On tasks like legal research, first-draft contract preparation, document review and summarisation, due diligence checklists, and routine regulatory analysis, he says work that previously took a junior four hours is now being completed in under one hour via the latest AI.

“The work that’s being compressed is the bread and butter of junior and mid-level practice.” Non-equity partners are impacted up the ladder according to Subramaniam. He says this cohort traditionally supervised and quality-checked high-volume work done by juniors, but there is now less of this work as the human output from juniors slides.

“If the first draft coming through is already 80 per cent of the way there, the review and iteration loop shortens considerably. Fewer revision cycles means fewer recorded hours,” he says. The “counterintuitive but entirely logical flip side” of this trend is a heavier workload for senior associates and equity partners.

Subramaniam says as AI compresses the “execution layer” of legal work, more time and attention shifts upstream to advisory work, strategic judgment, client management, and complex problem-solving. These tasks remain outside the realm of AI. “Senior associates and equity partners are the ones absorbing that work,” he says.

“There’s also a supervision dynamic at play. Someone still needs to review and validate AI-assisted output, exercise judgment on the grey areas, and manage the client relationship, and that responsibility falls to senior practitioners.

“In many firms, equity partners are also spending more time on business development, firm strategy, and navigating the AI transition itself – all of which adds to their recorded hours.”

Morris Misel, a business futurist and strategist, echoes the comments, arguing that as AI compresses lower level tasks, workloads are tending to move upwards. Misel says work “complexity” is on the rise for senior associates and equity partners as clients are increasingly unwilling to pay for process-oriented work. They instead want “judgement, nuance, risk interpretation and strategic advice” from top brass.

The other driver, he says, is a concentration of accountability in senior ranks. “If AI produces a draft or a summary, someone senior still owns the outcome. The legal, reputational and commercial risk hasn’t disappeared – it’s actually heightened because the work is moving faster,” Misel says.

That means senior lawyers are reviewing more AI-generated work, making more final calls engaging earlier in matters that used to be handled lower down and spending more time with clients who expect higher-value input.

“In simple terms, the work hasn’t gone away. It’s moved up the chain.” The shift is not uniform across the sector, he says, and is more pronounced at the “big end of town” where AI take-up has been fastest.

The large firms also get more client pressure to demonstrate efficiency and typically have greater willingness to design workflows – another key driver. “In terms of practice areas, highly document-heavy fields like M&A, banking, and large-scale litigation are feeling it first,” Misel adds.

“Areas requiring deep human nuance like family law, criminal law or complex negotiations are changing more slowly, but they will still be affected. “What’s interesting is that the more structured the work, the more exposed it is to AI compression.”

Katie Richards, the chief executive of online legal firm Virtual Legal, agrees, saying the big firms have budget for better systems and to employ testers on AI. These systems can then be rolled out across entire workplaces, shifting work patterns.

Whereas “the smaller end of town has less resources but can move faster in implementation”, Richards says.

“Most of the AI tools have been in the litigation & M&A space for eDiscovery but contract drafting is now getting better for the commercial guys.”

Richards says with AI speeding up work done by juniors, more jobs can come in the door flowing through to a rise in non-AI work done by firm leaders, such as face-to-face meetings.

“The more senior roles are using analytical thinking beyond the AI capability and spending time in meetings with the clients.”

She cautions that the increasing use of AI by young staff comes with risks. These include a lack of thought about client privilege and privacy considerations and putting client data into systems that are not secure.

“One firm I know of recently had a recording of a QCAT hearing and at the end the AI note taker emailed a copy to QCAT too, so that firm was hauled in front of the Legal Services Commission to explain their actions of recording the meeting without consent.”

According to futurist Misel, the redistribution of work prompted by AI will accelerate in years ahead “not because the tech will suddenly become perfect but because the economics demand it”.

“Clients now know what’s possible,” he says. “Once a client sees that something can be done in hours instead of days, they won’t go back to paying for the old model.” This will have flow-on effects to firm financials, he says, arguing that the more efficient firms become with AI, the harder it becomes to justify traditional billing models to clients.

He says it’s likely more clients will put the spotlight on billable hours. “The deeper signal is that billable hours as a concept start to come under pressure,” he says.

“It won’t  disappear overnight, but weaken as the primary measure of value.

“We’re moving, slowly but clearly, toward outcome-based and value-based pricing in parts of the legal sector.”

Research from legal talent platform Axiom, which entered the Australian market in 2022, suggests firms are not yet feeling pressure to drop their rates for AI-assisted work. According to a recent company survey of more than 600 senior legal leaders globally, 79 per cent of respondents are actively using AI tools in their practices.

“Yet instead of passing the substantial efficiency gains to their clients, the vast majority are pocketing savings – and in many cases, charging even more for AI-enhanced work,” it says.

The survey found just 6 per cent of law firms charge less for AI-assisted work, with 34 per cent of law firms charging more for AI-enhanced services than traditional legal work. Some 58 per cent of law firms have not cut their rates despite AI assistance, it found.

Syed Asif Ali, founder of Point Media, a digital consultancy that works with APAC law firms, is less strident about the impact of AI inside firms. He says it is not always true that redistribution of work due to AI means less work for juniors.

Ali, who has been working with an undisclosed mid-tier firm in Sydney, says many juniors are now spending significant time deploying tools like Lexis+ AI and Westlaw Precision for research and Kira or Luminance, for contracts and due diligence work.

By contrast, “senior associates and partners are now spending more time reviewing and validating AI-generated output instead of doing the work from scratch,” he says.

“That sounds efficient, but in reality it creates a different kind of pressure – you’re no longer just doing the work, you’re responsible for catching what the AI might miss.”

The comments fit with a recent Citi-Hildebrandt annual client advisory survey, which included responses from many large and mid-size US-based firms, said 63 per cent of big law firms tipped GenAI to lead to some change in the lawyer leverage model by 2035.

Ali points to another shift, related to mid-career lawyers, whom he says are increasingly falling between the high-level tasks of senior staff and the AI review work by juniors. As he puts it: “Mid-level lawyers seem to be the most affected here”.

“They’re too senior for repetitive tasks, but not always positioned as final decision-makers. that creates a bit of a squeeze in the middle.”

In his view, the rebalancing of law firm workloads is also having unintended consequences. For instance, he says he has been told that juniors, who before AI learned about their practice area by doing tasks like document review, are now learning by fixing AI mistakes.

“It’s a completely different kind of training,” Ali says.

Like Misel, Ali highlights the impact of changing workloads on billing models as AI usage lifts. “The billable hour model starts to break,” he says. “Clients are increasingly questioning why they should pay full rates for work that is partially automated, while firms are still figuring out how to price AI-assisted services.

“The biggest challenge isn’t the technology itself, it’s the adjustment in how work is structured, priced, and trusted.”

University of NSW academic Michael Legg has identified a similar trend in what he calls the “flattening pyramid” of the Australian law firm. Legg, a law professor, likens the traditional structure of a law firm as a pyramid, with many junior lawyers down the bottom.

“They skill up, and they either go up through the ranks to become partners, or they’ll go out and take on different roles,” he says in a recent note.

“However, with AI taking over a lot more routine types of work, there will be significant automation of the work that junior lawyers traditionally undertook.

“While it’s unlikely that work will be fully automated, the impact on large cohorts of junior lawyers and paralegals will be significant.” Even so, his prediction is that a large number of hours will still be worked by juniors – it’s just that the nature of the work will change.

There will be emerging roles at the juncture of legal expertise and technological capability to somewhat offset the loss of traditional roles, according to Legg, director of the university’s Centre for the Future of the Legal Profession.

He says: “Firms will require law tech solutions advisers and legal data scientists who can bridge the gap between AI-generated insights and legal judgment”.

“You’ll need your law tech solutions advisor or your legal data scientist, who’s able to help lawyers by collating data, producing an analysis and generating insights for lawyers.”

Against this backdrop, hiring priorities are rapidly changing to accommodate AI. Australian law firms are pulling back from investing in junior staff and instead making more lateral hires of experienced lawyers, echoing industry shifts in the US.

Indeed, a recent report found 86 per cent of large firms are planning to grow their ranks of associates overall through 2027. At the same time, just 35 per cent of firms planned to boost first-year lawyer classes, and only 37 per cent were onboarding more summer associates. Instead, firms planned to fatten senior ranks.

Legal recruiters have also witnessed more demand for lawyers with technological expertise and less demand for juniors without strong tech backgrounds.

Zed Law’s Nandan doesn’t see these changes slowing down anytime soon. “It will absolutely accelerate,” he says. “This is the year we’ll see the gap widen between firms that have embedded AI into their workflows and those that haven’t.”

He says the next big thing when it comes to how work is done is “the next generation of legal AI, being agentic AI that can execute multi-step workflows autonomously”.

“It is already in deployment,” he says. “We’re moving from AI as a drafting assistant to AI as a junior team member that can research, draft, cross-reference, and flag issues across an entire matter.”