It’s the billion-dollar question of law: if AI purports to make doing your work faster, how do you bill by the hour? Of course, confronting this question also requires assessing a reality often crankily muttered by many under their breath: but it doesn’t make my work faster, because AI hallucinates so much that I spend most of my time correcting errors!
Meanwhile, the pressure to lower legal fees has already gone public. According to the Financial Times, Wall Street banks, including Citigroup, Goldman Sachs and Morgan Stanley, have put BigLaw firms on notice that they expect AI use to translate to lower legal fees. This comes at a point when, for Citi’s law firm banking group, demand for lawyers’ time somehow, counterintuitively, rose by 4.2% in the first half of the year.
What’s happening here? And how do we respond?
We spoke to a global expert in Responsible AI, Max Scott from StratAlliance Global. An early practitioner in Responsible AI governance, Max draws on international experience with Microsoft’s Office of Responsible AI to advise Australian law firms on practical AI adoption, governance, and commercial strategy.
At Microsoft, his work centred on a core challenge: how to deploy AI safely, ethically, and cost-effectively while defining clear measures of business value and return on investment. He now helps organisations in highly regulated industries turn Responsible AI principles into practical operating models, governance frameworks, and employee policies. This includes protecting sensitive data, managing bias and risk, establishing clear accountability, and ensuring AI-supported decisions are transparent, explainable, and commercially as well as ethically defensible.
Max was recently tasked with answering law’s thorniest technological question by a major Australian firm: How should they price AI-enabled legal services?
The good news: The question was never about whether AI makes your work faster. It’s about what your client is paying for in the first place.
Distinguishing legal production from legal judgement
"There are two types of legal services," Max explains. “There is legal production, such as drafting, document review, first-pass research, and other structured tasks. Then there is legal judgment: assessing risk, determining strategy, advising the client, and taking professional responsibility for the answer.”
This distinction matters because AI affects the two differently. “AI is particularly effective at accelerating legal production,” Max says. “But clients are often paying for something more than production. They are paying for a lawyer’s judgment, experience, accountability, and ability to answer a difficult legal question under uncertainty.”
Faster production does not therefore require an automatic reduction in the price of the overall legal service. “If the judgment, risk, and accountability remain substantial, the value may remain substantial even where AI has reduced the time required for parts of the work.”
The pricing implications depend on the fee structure. On hourly matters, AI-driven time savings must be reflected through accurate time recording. On genuinely value-priced fixed-fee or outcome-based matters, the firm may retain some of the efficiency gain as margin, provided the fee reflects value, risk, and the outcome delivered rather than an outdated estimate of the hours the work used to require.
For high-volume, production-heavy work, AI creates a different opportunity. Firms can lower their cost of delivery, offer more competitive or productised pricing, and operate at scale while continuing to protect premium pricing for work in which human judgment, strategy, and professional accountability remain central.
What does this mean for law firms?
It means firms don't need to apologise for using AI, discount reflexively, or scramble to justify their fees every time the subject comes up. They need to be clear, with themselves and their clients, about which part of the work they're actually charging for.
Where AI has changed things, of course, is production.
"In a situation where you're working on ‘legal production’, AI can enable a large production of documentation at volume and at scale. This is where opportunities can be gained because efficiencies can be scaled."
When tackling pricing, Max advises that firms look at the margin, not the hourly rate. “Firms that have genuinely cut the cost of producing high-volume work can compete at lower cost to out-compete competitors for production-based work without ever touching the price of judgment-based advice.”
It plays out differently by practice area
This isn't a one-size answer.
"High volume, low margin areas are the places where there's the most opportunity for automation with AI,” Max noted, pointing to work like debt collection. “Using AI alongside automation in high-volume, low-margin areas of practice can allow for a higher margin than your competition. It follows that if you have that higher margin, you can increase your volume of work, which translates to doing better business and making more profit."
Family law, by contrast, is highly human-focused.
"You're talking about sensitive subjects, people’s personal lives and their livelihood,” Max says. “No one really wants to do that with a chatbot. They want to do that with a human being. The value of that human being and their time should be paramount."
For high-touch practice areas, frame the value of what a client is paying for early on.
“What the client is primarily paying for is your human judgement and your human oversight, so be upfront about this and show how this will work,” Max says.
Doing so also shows you understand both a client’s concerns around costs and how AI tools can be effectively used in legal work.
"Disclosure is always the best option," Max says. "Before your client has to ask about how you bill and use AI, make sure you're disclosing it to your clients. It's always important to lean into the trust aspects of these tools."
How does AI governance and adoption fit?
This same clarity applies to the broader anxiety around AI adoption. Max sees firms stuck between two extremes: "I'm just going to buy Harvey or Legora, and somehow it'll be fine," versus avoiding AI altogether, which carries its own cost.
"If you avoid AI, you're going to find other firms are benefiting from efficiencies and becoming more competitive, at a lower cost and higher volume."
However, adopting AI without guardrails poses significant risks, as many will know.
“Showing up to court and having opposing counsel identify what's clearly a fabricated citation in your submissions could have serious repercussions for your career,” Max notes.
How do you address these risks? Start with clear principles, not tools.
"Organisations that are highly successful in navigating this AI moment align their values with how they operationally relate to AI systems,” Max explains. “These become your Responsible AI principles. Think of this like a Constitution, which governs how you’re going to interact with AI and how you will prevent misuse. Once you develop your principles, you’ll then be able to identify and categorise your controls, and from there, train your people so you can operationalise safe, effective AI best practice firm-wide."
By taking a principles-first approach, you protect yourself against the whims of swift technological change, because you’re building a foundation which endures.
“Even as technology changes, your principles can remain the same,” Max says. “The principles function much like a constitution, from which power flows to govern how you’ll use AI. This means every time a new AI tool launches, you’re not stuck re-evaluating your whole governance process. You can choose to use a new AI tool within the framework of robust AI governance, grounded in principles aligned with your values. This is what Responsible AI means in practice, and it’s what provides the guardrails for safe AI use in many cutting-edge organisations.”
What does this mean for Australian firms?
Alison Laird, Director of the Centre for Legal Innovation (CLI) at The College of Law, says the questions Max is answering are the ones she's hearing most from firms right now.
“CLI’s benchmark research tells us clients increasingly expect AI-enabled efficiencies to flow through to fees, yet many firms are still working out how to measure those gains and translate them into new service and commercial models.
“The firms with the greatest advantage won’t be those with the most AI tools. They’ll be the firms that can explain what clients are paying for, where technology adds efficiency, and where human judgement, critical thinking and context create value.
“AI can generate an answer. The lawyer’s value is in understanding what that answer means in context. Clients don’t need a technical explanation of every tool in the stack. They need confidence that their lawyer knows when to use AI, when not to use it, and where human judgement remains essential. In this market, clarity builds trust.”
What’s clear is that the question of how to price AI-enabled legal services isn’t going away. If you haven’t already been asked, it’s unlikely you’ll be exempt much longer, given what’s already happened on Wall Street. The good news is that between Max’s framework and Alison’s own evidenced-based observations, you have a way to proceed: judgement is still judgement, and remains innately human. Legal production has simply gotten cheaper to deliver, and a firm that understands the difference is a firm that wins the trust and favour of clients in the long run.
![How to handle Direct Speech after Gan v Xie [2023] NSWCA 163](https://images4.cmp.optimizely.com/assets/Lawyer+Up+direct+speech+in+drafting+NSW+legislation+OCT232.jpg/Zz1hNDU4YzQyMjQzNzkxMWVmYjFlNGY2ODk3ZWMxNzE0Mw==)
