At a glance
- Company
- Gilberts Chartered AccountantsAudit and accountancy firm
- Industry
- Professional services
- Where and when
- United Kingdom. Published July 2025.
- Who produced the figures
- The company itselfThe figure and the account are the firm’s own, written by one of its partners for the audit faculty of the Institute of Chartered Accountants in England and Wales.
What they did
Gilberts is an audit and accountancy firm in the UK. One of its partners described how it took up AI in an article for the profession’s institute. He calls it “a series of small, practical steps”.
The firm began with the AI features in software it already used, and only on low-risk internal tasks. It now uses AI for:
- summarising long documents such as board minutes, lease agreements and loan terms;
- matching documents and comparing versions, which helps when reviewing accounts and auditing large numbers of invoices or contracts.
Around this it set a few rules:
- No client data in public AI. Nothing confidential or specific to a client goes into open AI tools. This is a firm-wide rule.
- Clients are asked. The firm talks to clients regularly to get their approval for the use of specific AI tools.
- Every output is a first draft. It gets the same scrutiny as the work of a junior member of the team.
- A short policy. Simple, concise and focused on the main principles.
- Leaders go first. Partners and managers use the tools and share their experience. Staff are regularly reminded about data protection, confidentiality and checking outputs.
The firm is also looking at closed AI systems that would run inside its own infrastructure.
The numbers
- Summarising a long document: 10–15 minutes, down from 45, with human review included.
What the case shows
- A smaller firm can start with what it has. It used tools it already had and chose tasks where a mistake would do little harm.
- The data rule came with the first use. The partner writes that clear rules on data matter most where there is no dedicated IT team or in-house lawyer.
- Existing duties carry over. Staff already understood confidentiality and professional ethics. The firm applied them to a new tool.
- Clients were told. The firm now explains its use of AI in first conversations with prospective clients.
What this says about the practices
Practices the company applied
- Start with ready-made AI in software you already use, before building your own The firm began “with tools the firm already has and use cases involving only low-risk internal tasks”. Do first · Effort: low (days, one person)
- Decide what data may go into which AI tools It is “a firm-wide rule that no client-specific or confidential data is to be entered into open/public AI models”, and the firm asks clients to approve specific AI tools. Do first · Effort: low (days, one person)
- Keep a person checking anything that goes to a customer or carries legal weight AI output is treated as a first draft and given “the same scrutiny as you would to a junior team member’s work”. Do first · Effort: medium (weeks, a few people)
- Write a one-page AI-use policy staff will actually read The firm’s advice from its own experience is to create a simple AI usage policy, keep it concise and focus on the key principles. Do first · Effort: low (days, one person)
- Protect psychological safety during an AI rollout Partners and managers use the tools and share what they find, so that staff “feel safe to try new tools but are clear on what is acceptable”. Do first · Effort: medium (weeks, a few people)
Sources
- Luke Parker, Partner, Gilberts Chartered Accountants (July 2025). AI: a smaller firm’s perspective. Audit & Beyond, ICAEW Audit and Assurance Faculty. icaew.com