How GCs are Rethinking Legal Work With AI
Three in-house leaders share how they’re expanding their teams’ capacity while keeping lawyers’ judgment central to the work.
In-house legal teams face a steady stream of work that varies in urgency, complexity, and the judgment required. AI can accelerate first drafts and other preparatory work, giving lawyers more room to focus on decisions that require their deep knowledge of the business. That raises an important question for legal leaders: Which work should the team take on vs. hand off to AI, and where can its lawyers add the most value?
At Harvey FORUM in Sydney, Silv Roest, Group Chief Legal, Corporate Affairs, and Sustainability Officer at The Warehouse Group; Andrew Chivers, Regional General Counsel, APAC at Brambles Limited; and Georgette Apostolopoulos, General Counsel at The Star shared what they’re learning as AI becomes part of their teams’ day-to-day work.
Their examples offer practical lessons for using AI to change how legal teams work, beyond accelerating individual tasks.
Change How the Work is Divided
At The Warehouse Group, one of the largest retailing groups in New Zealand, Roest is looking at how a four-lawyer team spends its time. First drafts, summaries, and issue tables are necessary work, but they can make it difficult to reach the questions that need the team’s judgment and knowledge of the business. That makes the division of work a useful place to start.
Apostolopoulos faced a different capacity challenge at The Star, one of Australia’s leading integrated resort companies. After the legal team downsized, it no longer had an in-house litigator to run an anti-discrimination matter. Her team used AI to help prepare the evidence before asking outside firms for quotes. By preparing the evidence in-house, the team could ask firms to price the work that remained, giving it a way to manage outside counsel costs.
Both examples start by breaking work into its component steps. Legal leaders can then identify what AI can help prepare, where their lawyers need to apply judgment, and when outside counsel’s expertise is needed. That gives a lean team a clearer way to allocate its time and scope external work.
Make Existing Knowledge Easier to Use
For a legal team supporting a global business, even a familiar question can take time to answer when the relevant guidance sits across different policies and jurisdictions. Global supply chain logistics company Brambles has 24 lawyers working across 60 countries. Chivers described how his team brought its policies together in an agent that people can query, reducing the need to search through documents each time a question comes in.
The example offers a useful starting point for teams with knowledge spread across many documents: look for questions that come up repeatedly and require lawyers to consult the same sources. Making those sources easier to use can help the team respond more quickly, while lawyers still assess how the guidance applies to the particular situation.
Help Lawyers Evaluate AI Output
Making information easier to access also changes the skills lawyers need to practice. Chivers noted that junior lawyers may be comfortable using AI while still developing the legal knowledge needed to challenge its answers. Familiarity with the tool is useful; the ability to spot a missing fact, weak inference, or incorrect conclusion takes training.
That training can happen in the course of real work. When reviewing an AI-assisted draft, a supervising lawyer can ask a junior colleague what they checked, where the answer came from, and why they agree with it. The discussion helps the junior lawyer develop judgment alongside their ability to use the tool.
Apostolopoulos described a related challenge when colleagues bring Legal an answer they generated with AI. Before assessing it, her team has to understand what they were trying to do. “It’s been really important for us as an in-house team to go back to the business and validate the question. Not the answer, but the question: What problem are you trying to solve?” Clarifying that request gives Legal the context it needs to assess the answer.
Look Beyond the Legal Team’s Workflow
The next opportunity may be larger than improving a task inside Legal. Roest described a change in how she thinks about AI: a contract review is one step in a procurement process that involves people and decisions across the business. Improving that step can help, but the full process determines whether work moves more effectively from request to outcome.
For GCs, that means looking at who needs information before and after the legal team’s involvement. Where does work wait for a handoff? Which questions come back repeatedly? And would changing the legal step alone solve the problem? Those conversations can help teams choose AI projects that address how the business works, rather than adding tools to isolated tasks.
The three leaders’ examples begin with immediate needs: evidence to prepare, policies to navigate, and answers to check. Starting there gives legal teams a way to learn what works. It can also reveal where a change to one task calls for a broader change in how lawyers, colleagues, and outside counsel work together.
For more on how in-house teams are rethinking how legal work gets done with AI, read RSGI’s latest report, The New Economics of In-House Legal.
