Only a year ago, the words artificial intelligence made many in the legal profession feel queasy about the future. Now, the narrative has shifted. Firms are talking less about disruption – what AI will do to us – and more about opportunity – what we can do with AI. Richard Hartshorne, Business Unit Director at Expleo, charts the change in mood and tone.
What a difference a year makes.
In October 2024, Expleo invited leaders from major UK law firms to a supper club in London. There was a tangible sense of trepidation around AI and its potential impact. Fears related to disruption, crisis, lack of knowledge and loss of control were shared openly round the table.
Scepticism was another palpable emotion. Despite the noise around AI – and generative AI in particular – there were still few actual use cases in the legal arena. More pressing IT problems lay in legacy tech debt, rather than adopting shiny new applications. “Our AI eyes are bigger than our bellies,” one participant admitted, summing up the gap between ambition and reality.
In this meeting of minds, the greatest perceived threat from AI was centred on the response of clients. If general councils developed proprietary solutions, would they need legal firms at all? And if not obsolete, would legal firms be forced into painful transformation that would reimagine their conventional delivery and billing models?
The logic made for unhappy reading. If clients could cut costs with AI, they would expect external partners to do the same. That pressure could spark a ‘burning platform’ for deep structural change that the legal sector has long anticipated but largely resisted.
In this scenario, firms would need to compete not just with each other and native AI newcomers – but also retain their value in the eyes of increasingly self-sufficient clients.
Twelve months is a long time in tech
A year on, in October 2025, Expleo organised a similar meeting – this time over breakfast – and the AI fare seemed to be far more palatable. The consternation around the future had shifted to a sense of optimism. From ‘what will AI do to us?’ the question was now ‘how can we make AI work for us?’. Rather than take our jobs, AI can help us work smarter.
So, what had changed? For a start, firms had not spent the last year with their heads in the sand. They had brought in specialist talent who understood AI and data science. Younger, digitally minded lawyers were encouraged to pilot use cases.
One firm has developed an education app to teach colleagues about the possibilities of AI. Another was piloting ‘gift cards’ that allowed juniors to claim chargeable time on AI experimentation.
Keep a human in the loop
Perhaps most importantly – given the previous discomfort around general councils taking work in house – law firms were identifying the limitations of AI, as well as its strengths.
GenAI has potential for speeding up the review, research and drafting processes, and clients are keen to see those cost savings passed on. However, there is still no artificial substitute for legal acumen. Given the need for legal interpretation and verification; human judgement remains critically important for protecting client interests.
Yes, the lawyer of the future will need to be tech savvy, but they will still need to be legal experts, first and foremost.
Besides, from a reputational point of view, what firm can afford to caught cutting corners with AI? The global headlines around Deloitte’s slip-up in Australia delivered a stark wake-up call for professional services.
Setting guardrails
A major theme of this year’s meeting was governance. Legal firms are proactively tackling the threats associated with AI, from data privacy and confidentiality to accuracy, bias and ethics. Developing robust AI governance – known as Responsible AI – has become a priority to ensure firms stay ahead of both regulation and also the pace of technological change.
One invitee called for joint-ownership of governance across business functions, rather than letting it fall to IT alone to manage the risk. There was an interesting discussion around the need to regulate so-called ‘shadow IT’ – when tech is developed outside of official protocols.
Would people try to circumvent such a policy? And if so, what are their personal preferences and drivers. Are firms missing out by failing to bring these innovations within the perimeter?
AI is already stealthing its way into everyday use, such as joining meetings to transcribe or record conversations or firing off meeting actions to people who are mentioned. Does this aid productivity or raise confidentiality risks?
Risk of inaction
Over the last year, the tone around AI risk has softened. Firms must absolutely manage the risks that come with developing AI use cases. But now the greater hazard lies in inertia: lack of knowledge, failure to form an opinion, reluctance to prepare and adapt, even negligence in the face of evolving regulations.
Confidence and awareness are powerful antidotes to fear and uncertainty. The hype of AI had long since subsided, replaced by a steely focus on the skills needed to make AI a competitive success.
Everyone agreed that AI will only grow in importance. Therefore, legal firms must embrace the change both from a technological and cultural standpoint.
Food for thought. Same time next year?
