Many company lawyers have recently been shaking off their reputation as reluctant adopters of technology. Some have even emerged as among the legal industry’s more enthusiastic embracers of AI. The main draw is its potential to automate tasks, saving time and money for their own teams and potentially for the wider business.
Four leaders of legal operations teams explain here how they are assessing and applying legal-tech to help their companies’ lawyers do better work faster.
Common themes include using AI to automate legal tasks, encourage innovation and cut costs — as well as the challenges of calculating return on investment.
The four were selected by RSGI, based on its research for the FT’s Innovative Lawyers series.
Accelerating Business
This article is one of a six-part series on legal-technology that aims to address fast-changing business needs. Read previous articles here·
Russell Davies
Head of legal and compliance operations — Dentsu
Advertising and marketing group Dentsu operates in about 120 countries. Its eight-person legal operations team supports approximately 280 legal, privacy, data protection and compliance professionals working in the many different companies that comprise the group, including hundreds of subsidiaries.
One of Dentsu’s main IT systems is a legal operating system. The legal operations team uses a platform from LawVu, a New Zealand software supplier to in-house departments, to help it track legal requests and store information on thousands of individual and disparate legal matters, including contracts.
“It’s been foundational to bring those individual legal requests into one place,” says Russell Davies, global head of legal and compliance operations at Dentsu.
The lawyers use it to reference contract data and for helping to negotiate favourable contract terms. As Davies puts it: “Do you want to spend a lot of time negotiating a . . . media buying contract when you’ve already done it thousands of times before?”
He says Dentsu lawyers are building hundreds of their own “mini software tools” for completing legal tasks.
One popular tool was built using software from legal AI company Harvey that helps Dentsu lawyers choose which new business in advertising or marketing the company will pitch for. The software quickly highlights any potential conflicts of interest such as whether the agency already supplies a competing company in the same industry.
“We are a whole bunch of advertising agencies, and legal teams are inundated with pitch requests [from Dentsu business units],” says Davies. “Can we respond to this pitch? Can we do non-compete checks in our . . . byzantine [network] of 40,000 contracts?”
By using AI tools, about half of Dentsu lawyers are saving between four and eight hours each week, the group has estimated, based on a survey this year of 250 lawyers.
Because some Dentsu lawyers are building their own AI tools to automate legal tasks, managing decentralised tech development can present challenges.
Sometimes lawyers may not follow technical best practice or they unwittingly develop the same tools, which is not an efficient use of time or resources.
“The downside of democratisation of . . . [AI] workflows is that we’ve got duplication, [which] defeats the whole purpose of what we’re trying to do around automation,” says Davies.
Lawyers building AI tools may “not necessarily be doing it in the most consistent [way] and applying best practice”, he adds — including how data and the tools are maintained. Some of the AI tools could become “redundant quite quickly” without sufficient oversight, he says.
Anna Golovsky
Executive manager, legal and company secretariat operations — IAG
Anna Golovsky has worked in law firms and company legal teams for more than 30 years.
At Australian insurer IAG, she leads a six-person legal operations team that supports about 100 in-house lawyers.
Using AI tools and integrating IT systems for managing legal work and documents — from assessing data breaches to reviewing legal bills — has lightened the administrative load on in-house lawyers. More than half the 1,500 legal matters dealt with during a six-month review were at least partly automated, says Golovsky.
“The [AI] tools are partially automating, or fully automating, a significant number of matters, so that matters don’t have to touch a lawyer,” she says.
AI is helping to curb corporate spending on external legal advice, she adds. The legal operations team is using billing and AI software from Morae and IP Author to scrutinise bills from its panel of law firms.
The software evaluates how much time the law firm is spending on the work, the seniority of the lawyers managing the claim, and how the fee compares to previous law-firm charges for insurance claims. The software will flag any bill deemed unexpectedly high, and the law firm will need to resubmit the invoice to IAG, says Golovsky.
IAG receives about 50,000 of these invoices each year, making it unfeasible for human lawyers to review each invoice.
In just over one year, the software has produced a return on investment of more than 800 per cent, calculated mainly according to the cost of the software and how much it has helped cut external legal bills, says Golovsky.
More AI projects and automation are planned, including the creation of a new “AI orchestration layer” that “sits between” AI tools and legal software that stores information such as documents, legal requests and spending data.
“There’s always going to be an amount of [legal] work that needs to be done by a human, but it will be humans augmented by significant AI capacity,” says Golovsky.
Sarah-Jane Davies
Head of legal technology and operations — ArcelorMittal
Sarah-Jane Davies faces a particular organisational and financial challenge in her role as ArcelorMittal’s head of legal technology and operations. The international steelmaker’s legal function does not have its own IT budget or IT support. Within that, her own two-person legal operations team supports about 160 legal and compliance staff worldwide.
“We do not have IT dedicated just [for] the legal team,” she says. “So, we are very much in a queue with the rest of the [business]. We have to spend a lot of time speaking with the business, almost sort of persuading them . . . of the routes that we want to take.”
Legal-tech priorities have included using AI-based software to automate common legal tasks. For example, the company is using software from Leah, a contract-focused AI platform previously called ContractPodAi. The software can speed up procurement by letting staff insert pre-approved clauses into contracts.
The legal operations team has also worked with law firm A&O Shearman to trial a digital repository for sharing and analysing litigation data with law firms. Davies plans to use the technology with other external legal advisers.
Her team is planning to develop some bespoke workflows, queries, or AI agents — AI that acts autonomously — to review information on legal matters, which would then potentially be repeatable.
Calculating payback on much of its AI spending for ArcelorMittal’s legal team is tricky, says Davies. However, she reckons that its lawyers probably each save between four and six hours a week by using AI.
Davies uses several metrics to assess AI’s value, including helpfulness in creating any competitive advantage, but she cautions that this is “incredibly difficult” to measure.
She adds that calculating payback and negotiating legal-tech budgets may be further complicated if suppliers switch to charging customers based on how much they use the product, rather than the more usual basis of number of users licensed to use the product.
Consumption-based pricing would be “highly challenging” because it is “so unpredictable”, says Davies. “We still work on annual budgets.”
Gordon McCue
Head of legal operations — Shell
Shell’s legal function is reviewing its role in the international oil company, including exploring new uses of AI, as part of a push to reduce spending by about 40 per cent within three years.
Gordon McCue, head of Shell’s 35-person legal operations team, is leading the programme, which began in 2025 as part of a company-wide cost-cutting drive and was agreed with the chief executive. Shell’s legal operations team is based mainly in London, Houston, Kraków and Kuala Lumpur. Team members specialise in a handful of areas, including training Shell’s lawyers in tech skills, helping lawyers manage large legal projects, and assisting on negotiating fees for external legal advice.
McCue wanted to get an overview of in-house legal operations, which in 2025 employed about 900 people, including compliance and ethics specialists, starting with reviewing their purpose and value to the business.
“We decided to strip it back to the bones,” says McCue. “Where [does in-house legal] support the businesses? Where are we putting our time as a legal team? And where do we think the most value is for the businesses themselves?”
As a result, savings on legal costs will come from job cuts, spending less on external legal advice, and by using AI.
The legal operations team is using AI software from Microsoft and Thomson Reuters to help its lawyers complete some tasks, such as drafting and reviewing contracts, summarising meetings and emails, and preparing presentations. McCue says AI tools have helped Shell lawyers complete tasks faster, with estimated time savings now averaging between 15 per cent and 20 per cent.
“The main lever [is] about reviewing when lawyers should be actually doing things and when it could be kind of automated or it could be self-service, and simplifying things,” he says.
The legal function is on track to hit the 40 per cent cost-saving target, but that still requires navigating certain challenges, he adds.
Some Shell lawyers are using Microsoft Copilot software to design their own AI agents. The agents can answer frequent questions from staff in the business, such as on legal aspects of procurement, says McCue. Other lawyers, however, require more training on best practice in using AI.
“Some of our lawyers are really embracing it and are really good at this and are setting it up,” says McCue. “Others . . . just aren’t as comfortable with the pace of the transition.”