Why AI is No Longer Optional in Legal Work
Two legal leaders share how AI moved from experimentation to an essential part of delivering legal services.
Not long ago, AI in law firms was still framed as a question: something to explore, pilot, or selectively adopt. That framing no longer holds.
In this discussion from Harvey FORUM in London, Emma Dowden (Chief Operating Officer at Burges Salmon) and David Wakeling (Partner and Global Head of AI Advisory Practice at A&O Shearman) describe how quickly the AI conversation has shifted from experimentation to expectation, and from isolated use cases to firm-wide transformation.
For Wakeling, the turning point came early, when his team first encountered GPT-4 in a legal context. The capabilities were immediately clear, but what mattered more was what came next: recognizing that this wasn’t a tool to trial at the margins, but a technology that would continue improving rapidly and needed to be taken seriously from day one.
Dowden now sees that same shift playing out at the operational level. AI is no longer something firms can position as optional or experimental; it’s affecting how legal services are delivered, how firms are run, and how clients evaluate their advisors.
What makes this moment distinct is how broadly AI cuts across the business. It’s not just changing legal workflows, it’s reshaping three core dimensions at once: the business model of law firms, the day-to-day practice of lawyers, and the expectations of clients. Increasingly, clients are driving adoption as much as internal teams, asking firms how they use AI, not if they use it.
Turning Momentum Into Change
This shift is forcing firms to confront challenges that go beyond technology. There are structural constraints — partnership models, annual profit cycles, and unclear ROI — that make long-term investment difficult. And there are human barriers as well, like hesitation from leaders who feel they don’t fully understand the technology and concern from practitioners about how it will affect their roles. The firms making progress, like Burges Salmon and A&O Shearman, are addressing both.
They are moving beyond fragmented experimentation toward more coordinated approaches, putting governance in place to prioritize what works, identifying internal champions to drive adoption, and investing in the skills needed to support AI at scale. That includes not just lawyers, but developers, data specialists, and multidisciplinary teams that can translate capability into real workflows.
At the same time, the definition of legal expertise is beginning to change. Increasingly, value comes not just from producing legal analysis, but from knowing how to work effectively with AI systems — how to guide them, evaluate them, and integrate them into client delivery.
In practice, AI hasn’t replaced core legal work, but it is changing how that work gets done and how firms compete. And both Dowden and Wakeling agree that choosing not to engage with AI is becoming harder to justify.
