AI Tools for Accountants #5 – Family Trust Distribution Tax (FTDT) Risk Tool

Any Australian accountant plugged-into the industry zeitgeist would have recently placed Family Trust Distributions Tax onto their ever-growing list of client niggles. Stories of the ATO raising FTDT assessments from years past raise doubts about the impact of previous Family Trust and Interposed Entity Elections, which were once considered safe and ordinary steps to solve client trust issues at the time.

AI cannot help your clients if family trust ‘distributions’ have already found their way to parties outside the ‘family group’. However, it can help identify existing problems and highlight current and future risk areas with a view to avoiding family trust distribution tax going forward. Here’s how we went about creating a Family Trust Distribution Tax (FTDT) Risk Tool…

A word on Privacy:

Our FTDT Risk Tool involves the disclosure of client trust and business information, details of elections, test individuals, distributions and loans from family trusts etc. We have opted out of storage and data training with our enterprise subscriptions to AI models. This means client information is not stored or used for training the model after analysis; the data (in this case, information relevant to the FTDT Risk Tool) is deleted, with only the risk assessment output retained.

Designing the Tool

Directing AI models at problems through custom platforms like Elfworks allows the AI model to interact with the user to obtain specific information relevant to the problem at hand. This feature is perfect (and crucial) for a problem such as FTDT, as:

1. The user needs to be prompted to disclose all information relevant to the problem at hand (including information they may not have thought relevant); and

2. The model needs this information to complete an effective risk assessment.

The tool we built interrogates the user regarding family trust elections which then prompts further questioning e.g. test individual, distributions, non-commercial loans, use of franking credits etc. the tool then does the same interrogation when the user discloses a past interposed entity election. Against the background risk built through disclosure of past FTEs, IEEs and past actions (distributions, non-commercial loans from trusts), the AI model then builds visuals of ‘family groups’ for each ‘family’ trust and trusts linked through IEEs and highlights those actions that will (or have) given rise to a risk of FTDT assessments.

Which model and language to use?

The same with our Client Structure Builder, we used the Mermaid programming language, a JavaScript based diagramming and charting tool that allows for consistent production of diagrams, linked with a detailed prompt for Gemini 2.5.

How long does the FTDT Risk Tool take to complete?

Once the Tool has interrogated the user, the output of Diagrams + Identified Risks takes 20-25 seconds to produce. The interrogation steps can take 1-2 minutes, depending on the complexity of client affairs.

If you would like full, free, no obligation trial of an AI platform designed from the bottom up for Australian accountants with a Client Structure Builder (and dozens of others), please contact us at info@elfworks.AI or phone me on 0418 902 440.

Thank you for reading. Next blog I’ll take a look at AI Ideas for Accountants #6 – Division 7A Risk Tool.

Please click here for Elfworks pricing information.