Reflections on LawTechUK’s Generative AI Conference: Key Takeaways for AI Adoption

Earlier today, I attended the LawtechUK Generative AI conference. While I don’t work directly with law firms, much of my AI adoption consultancy involves helping marketing services firms, which, like law firms, have historically been built on billable hours. Over some very poor sausage baps I discovered that the challenges and opportunities AI presents in these environments are strikingly similar.

Here are some more widely applicable insights gleaned over the morning;

Buy AND Build: A Hybrid Approach to AI Adoption

Elliot White , Head of Innovation & Legal Technology Operations, Addleshaw Goddard

White emphasised that the traditional dichotomy of “buy vs. build” is evolving into a strategy of “buy AND build.” His firm has developed a secure and intuitive AI playground, providing the necessary tools (basically, prompts) for their legal teams to navigate AI effectively but has then used that foundation to develop their own bespoke point solutions.

His approach to onboarding lawyers included:

• Getting ahead of fear by addressing concerns proactively.
• Establishing a working group to mitigate apprehensions.
• A phased approach: Educate, Assist, and Adopt.
• Teaching lawyers prompt engineering on an internal platform, building confidence before moving to more bespoke AI applications.

They have since cultivated internal AI evangelists, measuring usage and engagement. Interestingly, once the mince pies had been digested post-holiday AI usage returned to pre-holiday levels, suggesting sustained engagement.

Regarding the downward pressure on fees due to AI, (surely the elephant in the room for all services firms) White acknowledged that while there’s no simple answer, there is still plenty of untapped opportunity before it becomes a pressing concern. (!)

Little Wins, Big Impact

Dee Masters Discrimination, technology and employment Barrister

Masters’ message was clear: small AI-driven efficiencies can create enormous ripple effects. She saves five hours a week using AI—a personal benefit for work-life balance. However, for a judge, that same five-hour saving equates to an entire day of court sittings.

She described AI as an “enthusiastic paralegal,” capable of:

• Conducting initial research for opening and closing submissions.
• Summarising thousands of pages of witness statements and legal arguments.
• Performing real-time gap analysis on submissions.

Currently, judges write notes by hand or type them, making the speed of a trial dependent on how fast a judge can write. Given that trials often face year-long backlogs, AI-assisted efficiency could have a profound impact on access to justice.

AI is Inevitable: Lawyers Must Adapt

Sir Geoffrey Vos, Master of the Rolls

I'd like to point out that Vos is both the second most senior judge in the UK, and is also very bullish on the role of Gen AI in the law. This is in an industry not noted for its embrace of technology or indeed change. (e.g. Vos' role has existed in the British legal system since the 13th century!)

But as you'd expect Vos' enthusiasm is for very sensible and compelling reasons;

• Every industry lawyers serve is already using AI.
• Lawyers must be adept in AI, as they will be central to AI-related disputes.
• AI will save time and money, enabling faster and fairer legal resolutions.
• Most importantly, AI can enhance access to justice.

He dismissed fear-based examples, such as the infamous case of Steven A. Schwartz, the New York lawyer who submitted hallucinated AI-generated citations. Instead, Vos provided clear guidelines for responsible AI usage:

• Understand what AI does—and doesn’t—do before using it.
• Avoid inputting sensitive information into public AI models to maintain confidentiality.
• Always review AI-generated outputs. Lawyers remain accountable for their work product.

Ultimately, Vos argued that some decisions will inevitably be made by AI because human-led decisions will become too costly and slow. The real challenge is deciding which decisions should be entrusted to machines and which should remain human-led.

Practical AI Use Cases for Law Firms

Catherine Bamford CEO, BamLegal

Bamford provided a practical “starter-for-ten” list of AI use cases for law firms:

• Drafting documents: AI should refine human-created drafts based on templates or past examples—truly unique documents are rare.
• Witness statements: AI can translate notes into structured statements and highlight any gaps.
• Recording events: AI can draft board minutes and summarise key points.
• Negotiation support: AI can perform pre-execution contract reviews and test agreements against best practices.

All useful use cases that done correctly would speed up processes and improve outcomes. I can quickly think of analogous use cases in marketing that should be the 'starter pack' for any agency Account Director.

A Rights-Based Approach to AI in Justice

Sophia Adams Bhatti, Chair, JUSTICE Programme on AI in Justice

Bhatti highlighted a new rights-based framework proposed by JUSTICE, a UK cross-party law reform charity. While there is much in the report that is of import for lawyers and the wider community it was their calls for AI applications in the legal sector to be rooted in human rights standards rather than generic ethical principles that stood out for me. The useful cross discipline insight here was "don't reinvent frameworks when there are  well established and widely understood practices already in place.

Governance: The Missing Piece in AI Adoption

Rowena Rix Head of Innovation, Dentons

Speaking on a panel, Rix shared two critical adoption strategies for conservative industries like law (and frankly marketing)

1. The first was pretty straight forward and will be known to anyone building or deploying digital services which is to involve the lawyers (stakeholders) from the outset to get buy-in on new tools and processes. It sounds obvious, but is so often ignored or forgotten that it is worth restating.
2. The stroke of genius was that Rix employed an AI governance lead within the innovation team. In doing so she gave the firm's lawyers, and clients, the certainty that everything the team does is in compliance with the relevant regulations. As this is so often cited in regulated industries as a reason not to adopt (I'm looking at you pharma marketers), taking this off the table with an explicit hire seems at once both revolutionary and should now be table stakes for any large innovation team, not outsourcing it to legal or IT compliance. 

Final Thoughts

Many of these insights are directly applicable to other professional services firms, particularly those still clinging to a billable-hours model. The key lesson? AI isn't replacing expertise—it’s amplifying it. The sooner firms embrace this shift, the better positioned they’ll be for the future. The challenge is that this expertise tends to sit with senior practitioners. If we're amplifying that expertise with AI then we need less grads and interns. As these intakes drop and there is less 'graft' for the few that are hired to cut their teeth on, how do we develop talent as it comes up through the ranks, when the ranks have been disolved?

I left the morning with a hopeful conversation with Wendy Jephson CEO of the startup Let's Think. Let's Think is aiming to capture the expert thought processes of senior team members so that they can reflect on it and refine it, but also so that it can be shared with others in the firm. I can see how a senior partner might resist having their brains and years of experience harvested like this, but at the same time it offers a possible solution to how we help young practitioners in all kind of professions gain access to senior experience when AI does away with the apprenticeship model and the chance to learn at a master's elbow.