I had one of those slightly surreal moments recently.
Earlier this year, I wrote an article for IFSC.ie titled How Are Small and Medium-Sized Solicitor and Accountancy Firms Adopting AI to Streamline Day-to-Day Work?. It looked at how smaller professional firms were beginning to use AI, the opportunities it presented and the importance of confidentiality, proper controls and human judgement. Journalist Sorcha Corcoran came across the article, got in touch and asked if I’d be interested in speaking to her for a feature she was writing for the Law Society Gazette. So, a big thank you to IFSC.ie for publishing the original piece. Without that, the conversation with Sorcha might never have happened!
We spoke about the opportunities, the risks, what I was seeing among professional services clients, and the practical reality for smaller firms that know AI is important but may not have the time, resources or appetite for some enormous “AI transformation project”.
Fast forward a few months and Sorcha’s article appeared as the cover story in the October 2026 edition of the Law Society Gazette. You can read the full feature, Swimming with Sharks, on pages 22-25. [Download the October edition (PDF).]
And, much to my amusement, one of my comments from the interview ended up giving the feature its title.
“Swimming With Sharks.”
The phrase came from a point I made about unmanaged AI use:
“AI adoption isn’t about staff copying and pasting sensitive material to any brand of tool. It has to be embedded into properly approved systems with clear controls. Without those, you’re swimming with the sharks.”
Seeing that line splashed across the front cover certainly gave me a laugh. But underneath the Jaws reference is a serious point.
AI use is probably already happening
For many firms, the starting point isn’t deciding whether to “adopt AI”.
People are already using it.
They may be drafting emails, summarising documents, researching a topic, restructuring text or testing prompts in ChatGPT, Copilot, Gemini, Perplexity or another tool.
So the first question I would ask a firm isn’t:
“Which AI platform should we buy?”
It is:
“What are your people already using, and what are they putting into it?”
That matters enormously in a legal environment where confidentiality, personal information and professional obligations sit at the centre of everything.
The risk isn’t AI itself.
The risk is unmanaged AI.
Start small
One of the things I have seen repeatedly with new technology is that adoption becomes much easier when people can see an immediate, practical benefit.
Trying to introduce a grand AI strategy covering an entire practice from day one is likely to frighten half the room and exhaust the other half.
Pick one problem.
Pick one workflow.
Test it.
It might be drafting the first version of a routine letter. It might be summarising a long document. It could be producing an initial chronology or helping turn a complicated explanation into clearer language for a client.
In the Gazette article, I described AI as useful for reducing friction so that solicitors can spend more of their time on judgement, client care, strategy and outcomes.
That distinction matters.
The objective shouldn’t be to replace the professional.
It should be to remove some of the repetitive work around the professional.
AI can summarise. It shouldn’t decide.
AI is particularly useful as what I think of as a compression tool.
Long documents can be summarised. Chronologies can be produced. Large amounts of information can be organised into something easier to work with.
For a smaller practice dealing with several live matters at once, that can be very useful.
But there is a line that should remain extremely clear.
The AI is not making the decision.
As I said in the Gazette, the human element, professional judgement and safety remain crucial.
AI can help get you to the information faster.
The solicitor still has to decide what that information means.
Four practical steps for a firm getting started
The Gazette also included a short framework I suggested for any small legal practice that hasn’t formally embraced AI yet.
- Find out what staff are already using. Make it clear that confidential client information must not be entered into unapproved tools, and choose one low-risk workflow to pilot.
- Create a short AI policy. It does not need to be a 40-page document. Make the rules clear and keep human sign-off at the centre of the process.
- Train people properly. Staff need to understand hallucinations, confidentiality, verification and why AI-generated work still needs to be checked.
- Review the tools before using client data. Understand where the information goes, how it is processed and whether the platform is appropriate for the work being carried out.
None of this is especially glamorous.
That is probably a good thing.
Good AI adoption should feel less like a technology revolution and more like sensible business process improvement.
Don’t automate the humanity out of the relationship
One of the areas where AI can be genuinely helpful is communication.
It can improve structure, simplify language and help turn something complicated into a clearer first draft.
But I would be very wary of allowing raw AI-generated language to become the voice of a professional firm.
Clients don’t build relationships with ChatGPT.
They build relationships with people.
In the Gazette article I put it this way:
“You don’t want soulless responses; the humanity has to come across or client retention will suffer.”
That applies far beyond law firms.
Solicitors, accountants, consultants, financial advisers and other professional services businesses are selling knowledge, judgement and trust.
AI can support those things.
It should not flatten them.
The opportunity for smaller firms
There is a temptation to assume that larger organisations automatically have the advantage because they have bigger technology budgets.
I’m not convinced that is always true.
A smaller firm can often make decisions faster, test a workflow with a few people and adjust it without multiple layers of internal bureaucracy.
The key is not trying to follow every AI development or buy every shiny new tool.
Understand the business problem first.
Then ask whether AI can help solve it safely and usefully.
That is a much healthier place to start.
Thinking about AI in your firm?
I work with professional services businesses on digital strategy, workflows, communication and practical AI adoption.
That can include identifying useful AI applications, reviewing existing staff use, putting sensible controls around it, developing AI policies and helping firms work out where the technology can genuinely save time without compromising confidentiality, judgement or client relationships.
If your organisation knows it should be doing something with AI but is unsure where to begin, feel free to get in touch.
Sometimes the best place to start isn’t with the technology at all.
It is with one annoying problem that keeps eating somebody’s time.
If your firm knows AI could save time, improve workflows and make day-to-day work more efficient, but you are unsure where to start or how to put the right controls around it, talk to me. I work with professional services firms to identify practical AI opportunities, review existing use, develop sensible policies and help teams adopt AI in a way that supports confidentiality, professional judgement and client relationships. Get in touch at guy@guyfagan.com and we can have a conversation about where AI could make a real difference in your business.
ArtificialIntelligence #AIAdoption #LegalTech #LawFirms #ProfessionalServices #AIForLawFirms #DigitalStrategy #LegalInnovation #BusinessEfficiency #HumanOversight #DataConfidentiality #WorkflowAutomation #IrishBusiness #IrishLaw #AIConsulting
