AI: a smaller firm’s perspective

Artificial intelligence (AI) has the potential to help audit firms of all sizes to enhance audit quality, improve efficiency and offer more value to clients – without needing a dedicated tech team or deep pockets.

As highlighted during the faculty webinar AI in audit: opportunities and challenges, AI is levelling the playing field in ways that were unimaginable just a few years ago. 

At Gilberts Chartered Accountants we have embraced AI not as a revolution, but as a series of small, practical steps – beginning with tools the firm already has and use cases involving only low-risk internal tasks.

We started by exploring the AI capabilities embedded in our existing tools. This allowed the firm to experiment in a low-risk environment, without needing to invest in new systems or extensive training.

It helps with your decision making on this journey to be aware of some AI basics such as common terms of reference, AI classifications, types of AI model and other considerations – and ICAEW resources can help with this.

Beginning with what you already have may be the most accessible entry point for AI adoption in many smaller firms. At Gilberts, we’ve found these good practices are also helpful.

Good practices and use cases

Our top five tips for using AI are: 

1. Start with internal use cases: Use AI to streamline internal tasks such as meeting summaries, email drafting and document preparation, before applying it to client-facing work. 
2. Always review AI outputs: Treat AI-generated content as a first draft. Apply the same scrutiny as you would to a junior team member’s work. 
3. Protect client data: Never input confidential or client-specific information into open/publicly accessible versions of AI-enabled software tools and AI models. Where possible, use closed AI or enterprise-grade tools that prioritise data security and confidentiality (see Data protection, below.)
4. Encourage experimentation but set boundaries: Foster a culture where staff feel safe to try new tools but are clear on what is acceptable. 
5. Create a simple AI usage policy: Keep this concise and focus on the key principles of your AI usage to ensure clarity for staff.

Be patient. It takes time to understand how your firm can get the best from AI. Don’t expect an immediate flood of use cases. As your teams become more familiar with what’s possible, the ideas will begin to flow.

Use cases we’ve identified include:

• Summarising long documents: AI tools can summarise lengthy documents such as board minutes, lease agreements and loan covenants. What used to take 45 minutes now takes 10-15 minutes, with human review. 
• Meeting notes: AI-generated transcripts and summaries allow everyone to be fully present in meetings without worrying about note taking. 
• Research and content generation: AI can be a great starting point for generating ideas or structuring content, saving time compared to starting from scratch. 
• Extracting and matching data: AI tools help automate document matching and version comparisons, which is especially useful when reviewing accounts and auditing large volumes of invoices or contracts. 

These AI tools don’t replace auditors – they free them up to focus on key judgements, applying professional scepticism and client interactions.

Balancing efficiency with risk

While the benefits of AI for smaller firms are clear, it is not without risks.

1. Data protection

It is critical to establish clear policies around data usage, especially in the absence of dedicated IT teams and/or legal counsel. At Gilberts, we’ve made it a firm-wide rule that no client-specific or confidential data is to be entered into open/public AI models. We also ensure regular communication with clients to obtain their approval for the use of specific AI tools. 

This isn’t just about compliance with data protection regulations, it’s about maintaining client trust.

At Gilberts, we are also exploring closed AI environments: self-contained systems where AI models and tools will operate within the firm’s infrastructure to allow for more secure experimentation. As these tools become more widely and easily accessible, we expect more firms to adopt them to unlock AI’s potential without compromising data security.

2. Quality control

AI tools can produce outputs containing convincing but factually incorrect and nonsensical information – known as hallucinations – posing real risks in audit. Some tools are even designed so that their outputs appear to be correct even when they are not. 

For smaller firms, quality control starts with awareness: 

• Pilot AI on non-critical tasks and always check the sources are reliable. 
• Treat the output as you would treat a first draft review. Keep humans in the loop, especially for judgement-based decisions. 
• Carefully document inputs and outputs on file, as you would with any other logic that informed your decision making. 
• Most importantly, creating a space for open discussion concerning AI usage is key to safe, effective adoption.

People-led, AI-enabled

AI adoption is as much about people as it is about technology. That is why, at Gilberts, we have worked hard to create a culture of open dialogue and shared learning.

We encourage our team to talk about what is working, what isn’t, and what they’ve learned. These conversations help us to evolve our understanding and build confidence across the team.

Leadership plays a key role. When partners and managers use AI tools and share their experiences, it sends a powerful message: this is something we are all learning together. But we are equally clear about boundaries. We regularly remind staff about data protection, confidentiality and the importance of verifying outputs.

Key principles around professional ethics, client confidentiality and data protection regulations are already well understood by our teams – they just need to be applied in a new context. As we move toward more advanced tools, we will continue to reinforce these core principles.

Client expectations and advisory opportunities

While we’re not positioning ourselves as AI consultants at Gilberts, we are seeing increasing opportunities to open conversations and make referrals where appropriate.

We’re also starting to talk about our own use of AI during client onboarding. While we don’t receive many formal tender invitations like the larger firms, we do have fact-finding calls with prospective clients. These are great opportunities to explain how AI helps us deliver higher-quality audits, faster turnaround times, more partner involvement and a better client experience.

Looking ahead

The auditor of the future will need more than just technical knowledge. A new skill set is emerging: the ability to interact effectively with AI tools. This includes:

• Data literacy: Understanding how to interpret and interrogate data outputs. 
• Prompt engineering: Knowing how to ask the right questions to get useful responses. 
• Critical thinking: Evaluating AI outputs with professional scepticism and context.

Professional judgement, scepticism and ethics will remain at the core of what auditors do and the skills they need for this. It’s not about becoming coders, it’s about becoming confident users of powerful tools. As AI becomes more embedded in audit processes, we’ll increasingly need to document how we use these tools, so that others can review and understand the judgements made.

We believe the nature of audit will evolve and that AI tools are going to play a pivotal role in shaping this. 

For smaller firms, moving to fully digital systems can present challenges. Many clients still rely on legacy systems, paper records or non-integrated platforms, but as technology advances and tools evolve, new opportunities for tech-driven approaches will emerge.

Final thoughts

AI is not a silver bullet but it can be a powerful tool. For smaller firms, the key is to start small, stay curious and focus on quality. Let your teams experiment, but always protect your data and always review outputs as if they were a first draft. 

The future of audit will be shaped not just by the AI tools we use, but by how we use them. In that respect, smaller firms have a real opportunity to lead with agility, integrity and innovation.

Luke Parker, Partner, Gilberts Chartered Accountants