AI and the law: the time to wait and see is over

Artificial intelligence (AI) has already crept into the way many of us work, even if we don’t call it that. Software that drafts notes, searches documents or predicts what we’re typing next is AI.

People wonder whether it will change the legal profession, but it already has, just not everywhere all at once.

Some lawyers see it as progress, but others are wary. I understand both reactions. The law is built on caution and accuracy after all.

AI isn’t something waiting patiently at the edge of our profession; it’s already woven into it. The real question is how we choose to respond.

Where the law profession stands

Lawyers tend to be careful adopters of new technology, and that caution has served us well, but the pace of change is accelerating.

A recent LexisNexis survey showed that 61% of UK lawyers now use generative AI in their work, compared with 46% eight months previously. Clearly this isn’t hype, it’s a signal that the profession is quietly shifting.

Large firms are testing specialist AI systems while many smaller firms are experimenting with affordable, cloud-based tools that help them compete on speed and cost.

Across the sector, many firms that were in ‘wait and see’ mode are now beginning to engage more directly with AI.

It’s easy to view AI adoption as a technology project, but it’s really about how lawyers manage knowledge and time.

Every hour saved on research or admin can be spent on strategy or client work, and that’s the practical reason AI matters.

A lesson from dinner

At an event dinner in New York a few weeks ago, I was seated next to a former US CFTC Commissioner and senior attorney.

She mentioned that she’d tried one of the better-known AI tools and hadn’t been impressed. When asked a question about crypto-markets regulation and fraud, the answer she was given was only partially correct.

I pulled up my phone. Instead of relying on the tool’s default mode, I used a version connected to verified source material through retrieval-augmented generation (RAG).

After adding a bit of context to the prompt and phrasing the question as we might brief a junior colleague, the same tool came back with a complete and accurate answer, with references.

It wasn’t the software that had changed, only how it was used. That short exchange summed up the challenge we face as a profession.

AI is like any other instrument: the result depends on the skill and experience of the person using it. When handled well, it’s a powerful assistant; when not, it’s just another distraction.

This isn’t just about unquestioning belief in technology though. It’s about being deliberate with the prompts we use, to ensure it’s useful and gives more value at the end.

How AI is already helping

AI is quietly improving the parts of legal work that we’d rather do less of, such as:

• research and analysis: AI can comb through case law and legislation in seconds, producing a first draft or pointing out elements we might have missed. It shouldn’t replace judgement but gives us a head start
• document review: during due diligence or disclosure, it can flag clauses or anomalies, sparing hours of page-turning
• routine admin: from scheduling to data entry, automation can remove low-value tasks that drain time
• firm management: some firms now use AI to track billing patterns and forecast workloads. These small gains add up quickly

According to PwC’s Law Firm Survey 2025, the UK’s top 100 firms expect AI to reduce chargeable hours by around 16%.

AI will not replace lawyers, but it will reshape how we work if we apply it to the right parts of practice.

The open question is whether those hours will be reallocated to higher-value work – or simply disappear – if firms do not engage proactively with AI.

Moving from experiment to strategy

In many firms, people are still experimenting with AI on their own by trying tools in their spare time.

This kind of informal use can easily slip into ‘shadow AI’ – well-intentioned but unmanaged experimentation – which is why firms benefit from setting light-touch guidance early.

The key next step is to bring it into the firm’s overall plan for efficiency, risk and growth.

The basics are simple enough. Everyone should know what these systems can and can’t do, and where the pitfalls lie, from biased outputs to data exposure.

Be clear on purpose and decide what you’re trying to fix. It could be research, contract review or the flow of internal knowledge.

Start small, on a contained task, then review the results and learn from them.

According to Thomson Reuters Institute research, organisations with a defined AI strategy are almost twice as likely to see positive results for their business as those without one.

Ethics and the road ahead

Confidentiality, accuracy and independent judgement remain key to the foundation of legal practice.

Before using any AI tool, we have to understand where data goes, how it is stored and who can see it.

It’s also always worth remembering that anything produced by AI still sits under our name, so always needs to be reviewed before it leaves the firm.

That approach keeps standards high and reassures clients that quality is not being compromised for speed.

This is increasingly reflected in emerging regulatory and professional guidance, which expects lawyers and firms to use AI in a controlled, transparent and responsible way.

Handled thoughtfully, AI can free up lawyers to focus on what they do well – exercise judgement, advise clients and solve problems.

By engaging with AI early in an ethical and confident way, and by setting expectations for how it is used, we can ensure it strengthens, rather than disrupts, the legal profession.

AI is no longer a future prospect; it’s already part of how our profession operates.

The question now is not whether to engage, but how.