Using AI to Draft Engagement Letters: Prompts and Examples for UK Accountants

Writing engagement letters is one of those tasks that takes far longer than it should. The structure is largely the same every time, the core elements are broadly consistent across most engagements, and yet many practices still start from a blank document or an outdated template. AI can change that.

This article is part of Runbook's complete guide to AI prompts for UK accountants. It covers how to use AI tools to produce a first draft of an engagement letter quickly, what a well-drafted letter will typically need to cover, and the prompts and examples you can use right now. If you want to understand whether your practice is set up to use AI well more broadly, the free AI Readiness Scorecard gives you a personalised picture in under five minutes.

Why engagement letters take longer than they should

Ask any practice manager where time gets lost in onboarding a new client, and engagement letters will come up. Not because they are technically complex, but because the task sits in an awkward middle ground. It requires professional knowledge to get right, but the actual writing is largely formulaic. That combination makes it easy to procrastinate on, easy to rush, and easy to let an outdated template carry on being used long after it should have been refreshed.

The result, in many practices, is a document that goes out later than it should, contains scope language that no longer reflects how the firm actually works, and occasionally misses something entirely. Professional body guidance on engagement letters is clear: the letter must set out the scope of services, the basis of the relationship, and the respective responsibilities of the practice and the client. Getting that wrong creates real risk.

AI does not solve the problem of professional judgement. What it does solve is the blank page. For a straightforward engagement, the first draft can often be produced in under a minute. The qualified person still reviews it, adjusts it, and signs it off. But the starting point is a usable document rather than nothing.

What AI can and cannot do here

It is worth being specific about where AI helps and where it does not, because engagement letters are a higher-stakes application than routine correspondence.

What AI is good at in this context

AI tools handle the structural and drafting elements of engagement letter writing well. Given the right inputs, a tool like ChatGPT, Microsoft Copilot, or Claude can produce a draft that includes all the standard required sections, uses appropriate professional language, reflects the entity type and services specified, and is written in British English throughout. For a simple sole trader or limited company engagement covering annual accounts and tax, the draft output is often genuinely usable with relatively minor amendments.

AI is also useful for adapting a standard template to reflect a specific client's circumstances. If a new client has an unusual service mix, or if you want to add a specific carve-out or clarification, providing that context in the prompt produces a more tailored draft than editing a static template manually.

What AI cannot do

AI cannot apply professional judgement to the specific facts of a client relationship. It does not know what your firm's approved limitation of liability wording says, what your complaints procedure states, or what regulatory body you are regulated by. It does not know whether a particular client situation requires additional clauses or disclosures. And it will produce confident-sounding output regardless of whether the content is technically correct for your circumstances.

This means every AI-drafted engagement letter requires review by a qualified person against your firm's approved template before it is issued. The AI draft is the starting point, not the finished product.

What a well-drafted engagement letter will typically cover

Before constructing a prompt, it helps to be clear on the core elements a well-drafted engagement letter will usually need to cover. Requirements can vary depending on the engagement type, entity, and your own professional body's guidance, so treat this as a practical reference rather than a complete compliance checklist. Your firm's approved template, and where relevant your professional body's guidance, remains the definitive source.

For many UK accountancy engagements, the letter will typically need to address:

• Scope of services: a specific list of what the practice will do, and an explicit statement of what is excluded from the engagement
• Client responsibilities: what the client must provide and by when, including record-keeping obligations and the requirement to provide accurate and complete information
• Practice responsibilities: what the practice undertakes to do, and the standard to which work will be performed
• Fees and payment terms: how fees are calculated, when invoices will be raised, payment terms, and the practice's position on late payment
• Regulatory status and complaints procedure: a reference to the professional body the practice is regulated by (such as ICAEW, ACCA, AAT, CIOT, or other, as applicable) and a mention of the complaints procedure available to clients
• Data protection statement: confirmation of compliance with UK GDPR and a description of how client data will be used and protected
• Limitation of liability: the wording approved by your firm, insurer, professional body guidance, or legal adviser; do not ask AI to generate this clause from scratch
• Signature block: for both the practice and the client, with space for date

The scope section is where most engagement letters fall short. Vague scope language is one of the most common sources of client disputes and fee write-offs. AI can help produce specific, detailed scope language if given the right inputs, but only if you tell it exactly what services are and are not included.

Prompts and worked examples

The following prompts are designed to work with any major AI tool. Copy the prompt, replace the bracketed fields with your specific details, and paste into your AI tool of choice. Review all output before use.

Prompt 1: Standard engagement letter for a limited company

This prompt produces a full first draft covering the required elements for a limited company client. It is the most commonly needed version for practices with a mixed client base.

What the output should look like

A well-constructed response to this prompt will produce a full letter of typically 600 to 900 words, with numbered or clearly headed sections, appropriate professional language, and a signature block. The key things to check on review are: that the scope section matches exactly what your firm will actually do, that the limitation of liability wording reflects your approved terms rather than something the AI has invented, and that the regulatory status reference is accurate for your practice.

Prompt 2: Engagement letter for a sole trader

The sole trader version differs in scope, regulatory references, and client responsibilities. The Companies Act obligations drop away; the self-assessment and record-keeping requirements become more prominent.

Prompt 3: Adding or amending a specific service

When a client's service mix changes (they add payroll, or take on VAT for the first time) you do not need a new engagement letter from scratch. This prompt produces a supplementary letter or amendment covering the new service only.

The review checklist before you send

Once the AI has produced a draft, the review step is not optional. It is where the professional judgement happens. The following checklist covers the most important things to verify before any AI-drafted engagement letter goes out to a client.

• Scope accuracy: Does the services list exactly match what your firm will do for this client? Remove anything that is not included. Add anything the AI has missed.
• Exclusions are explicit: If bookkeeping, payroll, VAT, or advisory work are not included, say so. Vague exclusions are the main source of scope disputes.
• Limitation of liability: Replace any wording the AI has generated with your practice's approved limitation of liability clause. Do not ask AI to create a liability cap or limitation of liability clause from scratch. Use only the wording approved by your firm, insurer, professional body guidance, or legal adviser.
• Regulatory status: Confirm the reference to your regulatory body is accurate for your practice. Practices may be regulated by ICAEW, ACCA, AAT, CIOT, or another body; the letter should state this correctly and not default to a body that does not apply to your firm.
• Fees are correct: Check the fee figure, the invoicing frequency, and the payment terms against what was agreed with the client.
• GDPR statement: Check that the data protection statement reflects your practice's actual data processing position. If you are unsure whether your AI usage is GDPR-compliant, consult a qualified data protection adviser before issuing the letter.
• UK English throughout: AI tools occasionally slip into US spelling or phrasing. Check for "analyze", "honor", "program" and similar.
• No fabricated references: If the draft includes a reference to a specific piece of legislation or professional body guidance, verify that the reference is accurate. AI can produce plausible-sounding but incorrect citations.

The engagement letter typically arrives partway through a wider new client onboarding sequence that also includes a fee proposal, a professional clearance request to the outgoing accountant, and a welcome communication. For a practical guide to applying AI across the full onboarding process, including ready-to-use prompts for each document type, read the Runbook article on how to use AI to streamline new client onboarding in your practice.

Data protection: what to check before using AI for engagement letters

Engagement letters contain client personal data: names, addresses, and details of the services being provided. Before inputting any of that into an AI tool, your practice needs to have considered the UK GDPR position carefully.

Free consumer tiers of AI tools are generally not designed for processing client personal data in a professional services context. They may not give the practice the contractual controls, data processing terms, retention commitments, or governance assurances needed for UK GDPR compliance. In many cases, inputs on free tiers may be used to improve the underlying model.

Business-grade accounts may provide the contractual framework and controls needed for professional use, including data processing terms. ChatGPT Business or Enterprise, Microsoft 365 Copilot under the practice's Microsoft commercial tenant, and Anthropic Claude for Work may provide data processing terms and business controls, but your practice should verify the exact contractual position before use. Specifically, you will need to confirm:

• the exact product and tier being used;
• whether a data processing agreement applies and what it covers;
• data retention settings and whether inputs are used for model training;
• international data transfer terms and where data is processed;
• security controls and certifications;
• your practice's internal approval and policy position on AI tool usage;
• whether client data should be anonymised or excluded from prompts in any case.

A business-grade subscription does not automatically make processing client data GDPR-compliant. The practice still needs to have the right internal policies, informed usage practices, and qualified advice in place. Take qualified data protection advice before processing client information through any AI tool.

If your practice has not yet confirmed its GDPR position on AI, the simplest interim approach is to work with anonymised or placeholder data in your prompts and populate the client-specific details once the draft is returned. This removes the personal data question from the drafting step while still giving you the time saving.

For a broader look at how to build your prompt approach across the full range of accountancy tasks, including data protection considerations for each category, the complete guide to AI prompts for UK accountants covers the practical detail.

Frequently asked questions