AI Essentials for Accountants

Disclaimer: All company names and scenarios used in this course are fictitious and created for illustration and training purposes only. Any resemblance to real businesses or organisations is coincidental.

It’s March, and your practice is in full swing. You’ve just finished processing the latest batch of GST returns for a group of small businesses, but one client is confused about their provisional tax obligations. They’ve received an IRD notice, and you need to explain it clearly-but your inbox is overflowing with emails, and you’re running short on time. You know the rules: provisional tax under the standard method is due in three instalments, and the Income Tax Act 2007 requires you to ensure clients meet their obligations. But how do you craft a response that’s both accurate and easy to understand?

This is a scenario many New Zealand accountants face during the busy season. With deadlines looming, clients needing clarity, and software like Xero or MYOB handling much of the data entry, the pressure to communicate complex tax concepts in simple terms is high. This is where AI tools like Microsoft Copilot can be a valuable ally-not by replacing your expertise, but by helping you streamline repetitive tasks and focus on what matters: your clients.

Generative AI is a type of artificial intelligence that creates text, images, or code based on prompts. Microsoft Copilot and similar large language models (LLMs) are trained on vast amounts of data to generate responses that mimic human writing. When you ask AI a question, it doesn’t “think” in the way humans do-but it uses patterns in its training data to produce answers that are often helpful, though not always perfect.

How AI Can Help in Accounting

For accountants, AI tools can assist with tasks that involve repetitive writing, data interpretation, or basic analysis. For example:

• Drafting emails to clients about tax deadlines, changes in legislation, or compliance requirements.
• Generating summaries of complex tax rulings or IRD notices.
• Identifying patterns in financial data (e.g., spotting irregularities in GST claims).
• Creating templates for common tasks like client onboarding, AML/CFT compliance checks, or tax planning documents.

However, AI has clear limitations. It cannot:

• Replace professional judgment or legal advice.
• Access real-time data from systems like Xero or the Companies Office.
• Make decisions about client-specific scenarios (e.g., whether to recommend a tax strategy).

Why This Matters Now

New Zealand’s regulatory landscape is evolving rapidly. The Tax Administration Act 1994 and Privacy Act 2020 place strict requirements on how accountants handle client data, while the Anti-Money Laundering and Countering Financing of Terrorism Act 2009 mandates robust compliance checks. AI can help you stay on top of these obligations by:

• Automating routine compliance tasks (e.g., generating AML/CFT checklists).
• Reducing errors in repetitive tasks (e.g., data entry or report formatting).
• Freeing up time for higher-value work, such as strategic tax planning or client advisory.

But it’s crucial to remember that AI is a tool, not a replacement for your expertise. Always verify AI-generated outputs against your knowledge of the law and your client’s circumstances.

Example Prompts: Ready-to-Use AI Commands

Here are three prompts you can use in Microsoft Copilot today:

Prompt 1: Draft a client email explaining provisional tax obligations
“I need to write a clear, professional email to a client who hasn’t paid their provisional tax. The email should explain what provisional tax is, the due dates under the Income Tax Act 2007, and the consequences of late payment. Keep the tone friendly but firm.”

Prompt 2: Summarise an IRD notice about GST changes
“Please summarise the key points of this IRD notice regarding changes to GST rules for 2024. Highlight the impact on small businesses and any deadlines they need to be aware of.”

Prompt 3: Create a checklist for AML/CFT compliance
“Generate a checklist for an accounting practice to ensure compliance with the Anti-Money Laundering and Countering Financing of Terrorism Act 2009. Include steps for client onboarding, transaction monitoring, and record-keeping.”

These prompts are designed to be specific and actionable, helping AI generate useful outputs quickly.

Common Pitfalls and Misconceptions

While AI can be a powerful tool, there are a few common pitfalls to avoid:

1. Over-reliance on AI-generated content
   AI can make errors, especially when dealing with nuanced legal or tax concepts. Always double-check AI outputs against your knowledge of the law and your client’s situation.

2. Assuming AI can access real-time data
   AI tools do not directly connect to your Xero or MYOB systems by default. If you need platform data, export only the minimum required fields and de-identify sensitive information before analysis.

3. Ignoring privacy and compliance risks
   Using AI to process client data must comply with the Privacy Act 2020. Ensure you’re not sharing sensitive information with AI tools unless you’ve implemented proper safeguards.

Try This: Draft an Email with AI

Exercise: Use Microsoft Copilot to draft a client email explaining provisional tax.

Steps:

1. Open Microsoft Copilot and use the following prompt:
   “Write a professional email to a client explaining their provisional tax obligations. Mention the two instalment dates (30 June and 31 October), the Income Tax Act 2007 requirements, and the consequences of late payment. Keep the tone clear and helpful.”
2. Review the response and adjust it to match your firm’s communication style.
3. Send the email to a client or save it as a template for future use.

This exercise takes just a few minutes and can save hours of repetitive work over time.

Key Takeaway

AI tools like Microsoft Copilot can help accountants save time on repetitive tasks, improve client communication, and support compliance with New Zealand’s regulatory requirements. However, they are not a substitute for professional judgment or legal expertise. By using AI strategically-such as to draft emails, generate compliance checklists, or summarise complex tax rulings-you can focus more on high-value work and less on administrative tasks. Start small, test prompts with de-identified inputs, and always verify results before sharing them with clients.