Responsible AI: how we build and use it

What does responsible AI mean for us?

For us, responsible AI means AI that a professional firm can explain, check and switch off. We design every system so that a person stays accountable for what it does, its actions are recorded, the people it deals with know it is AI, and your firm’s confidential information is handled with the same care as the rest of your practice.

This statement covers two things: how we build AI for client firms, and how we use AI in our own work, including on this website.

Who approves what AI does?

A person at your firm approves anything before it leaves your firm. Our AI agents and automations prepare emails, reminders, file notes and first drafts, then wait for someone with the right role to review and approve them. This is often called human in the loop.

The After-hours Receptionist is the one exception, because a live phone call cannot wait for approval. Instead, your firm approves the receptionist’s answer templates in advance, and it takes messages or books appointments for anything outside them.

How can you see what AI has done?

Every agent we build keeps an audit log: a record of what it read, what it prepared, who approved it and what was sent. Your firm can review the log at any time, and it is part of the handover documentation we provide, so you do not depend on us to find out what a system did.

Where we run a system for you under Managed AI, we review agent activity as part of the monthly service.

Will people know they are dealing with AI?

Yes. When an agent speaks or writes to people outside your firm, such as callers to the After-hours Receptionist or visitors using a website chat, it says that it is an AI. Messages that a person at your firm has approved are sent from your firm, in the same way as any other correspondence your staff prepare.

Does AI replace professional advice?

No. The systems we build help your people find information, prepare first drafts and deal with routine admin. They do not give legal, tax or financial advice, and professional judgement stays with your lawyers, accountants and advisers.

AI can produce confident but wrong answers, known as hallucinations. That is why AI output is reviewed by a person before anyone relies on it, why the Drafting Assistant flags every citation for checking, and why Private AI shows the source documents behind its answers. Our guide on whether AI is safe for Australian law firms covers the professional obligations involved.

How do we handle your firm’s data?

We treat your firm’s data as confidential and give AI systems access only to the information a task needs. Connections to practice systems are read-only wherever the task allows, and staff see only what their role permits.

When we set up business AI plans through AI Setup, we switch on the training opt-outs the provider offers and show your firm the settings. For the systems we build, our aim is to use AI services whose business terms exclude your data from model training, and we show you those terms before the build starts. Our security and data residency page explains where data is kept and how it is protected.

Do we make automated decisions about people?

No. We do not use AI to make decisions about individuals that significantly affect their rights or interests, and the systems we build are designed to prepare information for a person to decide. Our privacy policy explains how this relates to the Privacy Act’s automated decision-making rules.

Which guidance do we work with?

We use two Australian Government references when we design and review AI systems: Australia’s AI Ethics Principles, published by the Department of Industry, Science and Resources, and the National AI Centre’s Guidance for AI Adoption. Both are voluntary. We use them as reference points, not certifications, and we do not claim formal compliance with either.

Your firm’s own rules also apply, such as the professional conduct rules for lawyers and the Tax Practitioners Board’s guidance for tax agents, and we design each system so your firm can meet them.

Do we use AI in our own work?

Yes. We use AI tools to help research, draft and edit our own content, including this website, and to help write and test the software we build. A person at Pylon Digital reviews and takes responsibility for everything we publish or deliver. We do not put a client’s confidential information into AI tools that are not approved for it.

How can you raise a concern?

If you have a question or concern about how we use AI, or about an AI system we have built, please tell us through our contact form and we will reply by email. We review this statement at least once a year and whenever our approach changes.

Last reviewed 24 September 2026.