Tax season, automated: an AI checklist for accounting practices

Most of the strain in tax season is not tax. It is chasing documents, answering the same questions, and re-keying numbers. That is exactly the work AI handles well, as long as client data stays in approved systems and a registered agent signs off the advice. Use this checklist to find your first workflow.

Before the season starts

• Client checklists. AI builds a personalised document list for each client from last year's return and current notes, ready to send in your template.
• Engagement and onboarding. AI drafts engagement letters and prepares onboarding packs for new clients for a person to review and send.

While documents come in

• Document chasing. AI tracks what each client has sent against their checklist and sends polite, scheduled reminders for what's missing.
• Inbox triage. AI sorts incoming email into client documents, ATO correspondence, questions and admin, and files attachments to the right client folder.
• Data extraction. AI reads receipts, statements and summaries and fills a first draft of the workpapers, with each figure linked back to its source document.

While you prepare returns

• Client questions. AI drafts answers to common questions ("when will my return be done?", "what else do you need?") from the client's own file, for a staff member to approve.
• Status updates. AI sends short progress updates, so clients stop calling to ask.
• Review notes. AI lists anomalies against last year (a missing income source, a big jump in a deduction) so the reviewer starts with the right questions.

Controls that matter

• Approved systems only. Keep client data inside your practice software and an AI service your firm has approved. Do not paste client files into consumer chat apps.
• A registered agent signs off. AI drafts; your people decide. Treat every AI output as a first draft.
• Tell clients how you use it. From 10 December 2026, organisations covered by the Privacy Act must describe in their privacy policy any automated decisions that significantly affect people. Even where that does not apply, a plain sentence about how you use AI builds trust.
• Keep an audit trail. Record what the AI drafted, who approved it and when.

Where to start

Document chasing is usually the best first workflow: high volume, low risk, and every hour saved goes back into billable work. Build it inside the tools you already use, run it on a small group of clients, then roll it out.

Want to see which of these fits your practice? Ask Jarvis, our AI agent, at the bottom of our home page.

Source for the privacy requirement: Privacy and Other Legislation Amendment Act 2024 (new APP 1.7–1.9, in effect from 10 December 2026).