AI for Tax Agents: TPB Guidance, Risks and Next Steps

/ Accounting, Business, Other, Payroll, Rulings & Regulations, Taxation, Tips & Tricks
Applied Education Blogs

Published: April 2026

AI for Tax Agents: How to Use AI Without Falling Short of Your TPB Obligations

Last updated 17 April 2026

 

The Tax Practitioners Board’s draft guidance released in March 2026 explains how existing professional obligations apply when AI is used in tax agent services. Aimed at registered tax agents and BAS agents, it makes clear that while AI may improve efficiency and support client service, it does not replace professional judgement, accountability or the obligation to protect client information. It also indicates that client permission may be needed before client information is entered into AI tools, depending on how those tools are configured and used.

AI is moving quickly into tax practice. It can save time, streamline research, and support administrative work. For registered tax practitioners, the real question is not whether AI is useful, but whether it is used safely, ethically and in line with professional obligations.

For tax agents, BAS agents and practice teams across Australia, now is the time to review how AI is being used in your business. Not in theory, but in everyday tasks like drafting advice, summarising records, reviewing transactions, preparing forms and handling client communications.

Table of Contents

What has the TPB actually said about AI for tax agents?

The TPB has released draft guidance explaining how existing professional obligations apply when AI is used in tax agent services. Consultation closes on 21 April 2026, making this a current compliance priority.

The draft guidance says AI may help improve productivity, efficiency and client service. At the same time, it stresses that tax practitioners are still ultimately responsible for the services they provide, whether AI is used or not.

Who does the TPB AI draft apply to?

The TPB’s draft AI guidance is aimed at registered tax agents and BAS agents who use AI while providing tax agent services. It is also relevant to staff and contractors working within those practices, because AI use by team members still needs to be supervised and managed under the practitioner’s professional obligations.

What counts as AI in practice?

The TPB describes AI broadly and notes that, in tax practice, it can include:

  • analysis and auto-classification of transactions
  • research and interpretation of information
  • chatbot-style responses to specific questions
  • anomaly detection and error identification
  • auto-compiling tax forms using third-party information

Many practitioners are already using tools with these features, even if they do not identify them as AI.

Can tax agents rely on AI output?

No. AI can assist, but it should not replace a tax practitioner’s professional judgement. The TPB draft says practitioners remain accountable for the accuracy of all information and advice provided and should verify any AI-generated content before relying on it.

Why human judgement still matters

The TPB highlights a practical risk many professionals already recognise: AI tools, especially large language models, can be inaccurate, biased, incomplete or misleading, even when they appear credible.

For tax practitioners, that creates a clear boundary:

Use AI for support, not substitution

AI may be useful for:

  • drafting content
  • summarising source material
  • spotting patterns
  • preparing research starting points
  • organising data

But it should not be treated as the final authority on:

  • client-specific advice
  • tax law application
  • factual accuracy
  • professional conclusions
  • compliance decisions

These align with core obligations, including competence, maintaining relevant knowledge and skills, exercising reasonable care, and correctly applying taxation laws.

Practical takeaway

If AI helps prepare something that goes to a client, the review step is not optional. It is part of competent practice.

Do tax agents need client permission before using AI?

In some cases, yes. The TPB draft says practitioners must obtain client permission before disclosing information to a third party, and that may include entering client data into AI chatbots or copilots, depending on how those tools are configured and used.

Why is this a big issue? Many AI tools store or process data through third-party systems. If client information is entered, confidentiality obligations may be triggered. The TPB draft recommends clearly informing clients about: who receives the information, where it will be stored, and how permission is given.

Can I use ChatGPT, Microsoft Copilot or Claude with client data?

Potentially, but not by default. If a practitioner enters client information into tools such as ChatGPT, Microsoft Copilot or Claude, the TPB draft says this may amount to disclosing information to a third party, depending on how the tool is configured and used. In those cases, client permission may be required before that information is entered.

What good practice looks like

A safer approach may include:

  • Updating your engagement documents – Ensure engagement letters or consent forms clearly address AI use where relevant. The TPB notes permission may be captured through a signed engagement letter, signed consent, or another suitable communication.
  • Review the tool before use - Practitioners should assess commercial AI tools to check whether information will be kept secure and whether Privacy Act requirements are met.
  • Take extra care with TFNs - If the information includes tax file numbers, additional obligations may apply under the Privacy (Tax File Number) Rule 2015.

What compliance obligations should tax practitioners focus on now?

The draft guidance highlights five key areas: competence, reasonable care, supervision, quality management and confidentiality. These obligations are not new, but they apply equally when AI is used. Client information should not be pasted into any tool without considering consent, disclosure, security and privacy requirements.

What about tax file numbers (TFNs) and privacy?

If client information includes a tax file number, practitioners should take extra care. The OAIC states that the Privacy (Tax File Number) Rule 2015 regulates the collection, storage, use, disclosure, security and disposal of TFN information. In practice, that means practitioners should avoid assuming TFNs can be entered into AI tools in the same way as ordinary working notes or generic business data.

Before using AI with documents that may contain TFNs, consider whether the TFN can be removed, masked or excluded, whether the tool is appropriately secured, and whether any client permission or privacy disclosures are needed.

What should tax practitioners do now?

Treat this consultation period as a chance to tighten your processes before the guidance is finalised.

A practical checklist for firms and sole practitioners:

  1. Review where AI is already being used
  2. Identify what data goes into each tool
  3. Update internal policies
  4. Refresh client-facing documents
  5. Train staff on appropriate use
  6. Stay alert for the final guidance

Ready to build AI awareness without losing sight of compliance?

AI is changing how finance and tax professionals work. While the benefits are clear, so are the responsibilities. The focus should not be on avoiding AI, but on using it appropriately with the right processes and professional judgement.

At Applied Education, we support professionals stay current through flexible, online training designed for real working life. Explore our nationally recognised qualifications, skill sets and short courses in tax, bookkeeping, payroll, business and digital tools.

AI for tax agents

Stay current in a changing profession

Using AI well in tax and bookkeeping practice still depends on strong foundations in compliance, recordkeeping, reporting and professional judgement. Applied Education supports these capabilities through nationally recognised courses including:

Explore flexible online courses designed to help accounting, bookkeeping and payroll professionals build practical, job-ready skills for modern practice.

— The Applied Education team

Scroll to Top