For RICS members
ControlsIQ & the RICS AI standard
How ControlsIQ is designed to support RICS members in meeting their own obligations under the RICS professional standard Responsible use of artificial intelligence in surveying practice (1st edition, September 2025; effective 9 March 2026).
What the standard requires
The standard is a conduct standard: where it uses “must” the requirement is mandatory for members and RICS-regulated firms; “should” is recommended best practice. It applies to AI outputs that have a material impacton the delivery of a surveying service — for example, an analysis or summary you then rely on in a report or opinion. Construction-controls work in ControlsIQ (contract and claims analysis, notices, particular-conditions review, risk) will typically be in scope, so the duties below are the member's to meet.
How ControlsIQ supports your obligations
§4.3 — client communication / transparency
Your duty: Make clear, in writing, when and for what purpose AI is used in delivering the service.
How we help: A persistent, non-dismissable AI-use notice on every AI tool, and the AI Use Summary artefact (below) that a member can attach to a client deliverable to disclose what AI did, on what inputs, and when.
§4.2 — outputs, reliance and assurance
Your duty: Apply professional judgement to decide an output's reliability, and record that decision in writing under a named, responsible surveyor.
How we help: Every output carries a verification prompt directing the user to check it and have a qualified professional review it before reliance. The AI Use Summary includes a structured “Professional reviewer’s statement” the named surveyor completes to record the reliability decision.
§2 / §3.3 — baseline knowledge, risk management
Your duty: Understand the tool's limitations and failure modes; document AI-related risks.
How we help: Plain-language statements of what the tools are (probabilistic LLM assistance), what they can get wrong, and the vendor due-diligence information below — inputs a firm can fold into its own AI risk register.
§4.4 / §3.2 — explainability, AI-system register
Your duty: Provide, on request, written information on the AI system used, its working and limitations, and how it was assessed; keep a register of AI systems used.
How we help: This page plus a self-serve AI audit-log export (machine-readable record of every AI interaction: tool, timestamp, project, provider) the member can keep as evidence and feed into a firm register.
§3.1 — data governance
Your duty: Safeguard private and confidential data; understand training and retention.
How we help: Documented data posture: your content is not used to train any model, Anthropic's API retention is capped (then auto-deleted), data is encrypted in transit and at rest. See the Trust and Privacy pages and the sub-processor list.
§4.1 — procurement and due diligence
Your duty: Carry out and record written due diligence on a third-party AI supplier before procuring it.
How we help: The vendor due-diligence information below answers the standard's minimum written-request items in one place, so a firm can record its assessment.
The AI Use Summary artefact
From any project, a member can generate a one-page AI Use Summary — a printable record listing which AI tools were used, on what project, how many times and over what dates, the AI system involved, and a structured Professional reviewer's statement for the named, responsible surveyor to complete (the §4.2 reliability decision). Attach it to, or reference it in, your client deliverable to evidence transparent, reviewed AI use. A machine-readable AI audit logis available alongside it from Account Settings → Data & Export.
Vendor due-diligence information
Provided to help a firm record its written due-diligence assessment (§4.1) and answer explainability requests (§4.4):
Type of AI system
Large-language-model assistance via the Anthropic Claude API. ControlsIQ does not train its own models.
How it works / limitations
Outputs are generated probabilistically from the documents and data you supply. They are not deterministic, can contain errors, omissions or fabricated references, and are decision-support — not professional advice or expert evidence.
Training on your data
Contractually prohibited. Your content is not used to train ControlsIQ's, Anthropic's, or any other models.
Data retention
Anthropic retains API inputs/outputs for a short, capped window for safety purposes, then auto-deletes them. ControlsIQ stores extracted text on your project until you delete it or close your account.
Data & confidentiality laws
Processing is described in the Privacy Policy and KVKK Aydınlatma Notice; sub-processors are listed publicly.
Provider liability
Set out in the Terms of Service. ControlsIQ is decision-support software; professional responsibility for any deliverable stays with the user.
Practical testing
You can evaluate every tool on your own documents during use; the audit log records what was run for your records.
See also: Trust & Security, Privacy Policy, Sub-processors, Output Disclaimer and the Terms of Service.
Model-provider information (Anthropic), for items that sit with the provider rather than ControlsIQ: Trust Center, Commercial Terms (training prohibition, data handling), model documentation, and environmental-impact information. We can request specific written confirmations from the provider on your firm's behalf.
Firm-level oversight (Enterprise)
Several practice-management duties (§3.1–§3.3) sit at the firm level — staff training, an AI-system register, a quarterly-reviewed risk register, and oversight across a firm's people. These firm-level requirements are delivered as part of a ControlsIQ Enterprise engagement: a firm workspace with a compliance-officer oversight view across your members, the firm AI-system and AI-governance risk registers, and governance support scoped to your practice. Individual members can already download the §3.2 / §3.3 register starters and each member's AI Use Summary and audit-log export from Account Settings → Data & Export. Talk to us about an Enterprise engagement at controlsiq.com/enterprise.
Reference: RICS professional standard, Responsible use of artificial intelligence in surveying practice, 1st edition, September 2025 (effective 9 March 2026). ControlsIQ is published by Kozarda İnşaat Teknolojileri ve Danışmanlık Ltd. Şti. and is not affiliated with, approved, certified or endorsed by RICS. Nothing on this page is legal or professional advice, or a guarantee of compliance. Always work from the current version of the standard on rics.org.