Most of a surveyor’s time has gone into producing the report: finding records, reading them, measuring, pricing, assembling. The judgement clients actually rely on was fitted into whatever time was left.
AI changes that balance. It can take on the reading, the cross-checking and the first draft, at a scale no person can match, so professional time moves to where it matters: the building, the evidence that is missing, and the decision in front of you.
That is why Moyne Ross was built.
A building leaves traces everywhere: planning records, permits, imagery, listings, maintenance logs, past reports. Few assessments read them all, and fewer read them against each other.
The MR engine searches, reads and compares that evidence together, alongside any documents you share. Where sources agree, confidence rises. Where they conflict, the conflict is recorded rather than smoothed over. Every finding is held as documented, observed or inferred, so you can see what kind of claim it is.
More evidence, read more carefully, before anyone forms a view.
The engine was built around the real work of assessing a building, and it keeps each part of that work in its proper hands.
The model proposes. The platform calculates. The surveyor decides.
Each step leaves its working on the page, so it can be questioned afterwards. See what that looks like inside the Capital Register.
A capital figure is only as good as what sits behind it. In a Moyne Ross strategy, every priced line carries its cost basis: the scope assumed, the quantities, what is included and excluded, the date, and how firm the footing is.
Where a figure is a planning allowance rather than a quote, it says so. Where the evidence stops short, the register prices the inspection and goes no further. Confident language is not allowed to outrun the evidence, and the checks are built to catch it when it tries.
Honest uncertainty is more useful than false precision.
A first look at your building within hours, from its address alone, reviewed before you see it. No brief to write first.
Because the production is faster, the conversation can start earlier and go further. A different holding period, a refurbishment, a change of use: alternatives can be explored together within the agreed brief, with the register as the shared reference.
And the record stays live. New evidence, a quotation or a change of plan can be brought in and the strategy updated, rather than the work going stale in a document.
You also work with fewer layers. The person who built the engine is the person who reviews your building and talks it through with you.
AI does not inspect your building. It does not certify compliance, perform the final arithmetic, calculate the MR Index or take responsibility for the instruction.
A Capital Strategy issued as professional advice is reviewed and approved by the responsible chartered surveyor for its agreed purpose. A desktop assessment and an inspection-supported assessment rest on different evidence, and the strategy says which it is.
The engine extends judgement. It never replaces it.
RICS’s Responsible use of artificial intelligence in surveying practice standard took effect on 9 March 2026. It informs how Moyne Ross works: appropriate use, data and provider assessment, professional judgement, clear communication with clients, and accountability.
Moyne Ross is not currently an RICS-regulated firm, and this page does not claim RICS approval or certification of the engine. For any instruction, you can ask how material AI outputs were assessed, which assumptions affect their reliability, and whether further verification is needed. The extent of professional indemnity cover, including for AI-assisted work, is addressed before an instruction is accepted; this page does not itself confirm cover.
The practical arrangements behind the approach above. Open any section.
We use large language models to assist with research and analysis. They are general-purpose systems; Moyne Ross supplies the building methodology, evidence rules and output structure. The platform supports different providers, including Anthropic (Claude) and Google (Gemini), depending on the feature and selected configuration.
The provider used for a confidential instruction forms part of the processing arrangements explained to the client. A new provider is not authorised to receive private material merely because the platform can connect to it. Relevant changes are disclosed before the affected processing, with permission obtained where required.
Data retention and model-training terms differ by provider, product and account settings. We do not promise that information is retained only while a response is generated. Our Privacy & Cookie Policy explains this distinction and how to ask about the arrangements for your instruction.
An issued Capital Strategy records the professional advice for the agreed instruction. Where the portal offers interactive AI tools, such as chat or a generated working report, their status is different: those responses and reports may not have been individually reviewed or adopted as advice.
If a working output suggests a material new conclusion, ask us to review it and, where appropriate, incorporate it into the advice or a revised strategy. Generating a report does not by itself make its contents a newly issued professional opinion. The scope, permitted reliance and any further work are governed by your client agreement and instruction.
AI can produce convincing but incorrect statements, miss relevant evidence or give undue weight to a weak source. An apparently precise cost is still only as reliable as its scope, quantity, rate and assumptions. Our approach combines:
These measures reduce risk; they do not eliminate uncertainty or guarantee that every issue is identified. The speed of generation does not determine whether an assessment is suitable to rely on.
AI services may receive relevant asset documents, images, questions and assessment context, and may process them outside Australia. Private and confidential material requires an appropriate processing basis and express written permission before submission. We seek to minimise unnecessary personal information and address restrictions on third-party material before it is used.
The permissions for delivering an instruction are distinct from permission to publish a client case study or use identifiable material for model development. See our Privacy & Cookie Policy for providers, overseas processing, retention, research use and your choices.
You can ask which AI systems are involved, why they are used, what their relevant limitations are and how they affect your work. We will provide an appropriate written explanation, including relevant reliability reasoning and safeguards, while respecting security, confidentiality and third-party rights.
You can also contest a proposed use of AI, request correction or raise a complaint through our contact form. We aim to acknowledge concerns within five business days and give a substantive response within twenty business days, or explain when a response will follow. The complaints process in our Terms sets out the next steps.
AI is integral to an engine-generated Capital Strategy. A wholly AI-free version of that product is not offered. If you have restrictions on a provider or particular information, raise them before sharing the material so we can consider whether a different evidence basis or scope can meet your needs.
Updated 7 October 2026.
Start with an address. We’ll read the building before anything else is agreed.