FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Aipedia: Building control approvers
The Building Safety Regulator has no AI-specific rules, but its codes require RBIs and RBCAs to be trained in technology, manage its risks and remain accountable for every decision. Breaching a code can affect registration, and the same failures usually underpin negligence claims.
Part of: AI in building control approvers
In short
As at October 2026 the Building Safety Regulator has published nothing specific on AI in building control. The binding codes still reach AI use: RBI code 2.11 (technology training and risk), 3.1 (competence), 4.2 (accountability), 4.4 (no bias), 5.8 (records) and Annex 1.2 (auditable decisions); RBCA rule 2.19 (technology risk); and the operational standards rules on remote inspections and evidence. RICS members must also follow the mandatory RICS AI standard. A CROSS-UK safety panel has suggested approving authorities ask how far AI contributed to submissions.
The Building Safety Regulator registers building inspectors and building control approvers in England and acts as the building control authority for higher-risk buildings. It became a standalone body sponsored by MHCLG on 27 January 2026. Its three professional documents were first published on 7 July 2024 and last updated on 22 May 2026:
None mentions artificial intelligence, and the BSR has not published separate guidance or a statement on AI. The codes are binding on registered persons, so they are the starting point for any AI tool used in plans assessment, inspection or certification. Our AI in building control hub gives the wider picture.
| Provision | What it requires | How it applies to AI |
|---|---|---|
| RBI code 2.11 | Be trained and familiar with available technologies and manage their risks | Know what a tool can and cannot check before relying on it |
| RBCA rule 2.19 | Take steps to identify and mitigate risks of using relevant technology | A firm-level assessment and controls for each AI tool |
| RBI code 3.1 | Only carry out restricted activities you are registered and competent for | A tool cannot extend an inspector's competence or class |
| RBI code 4.2 | Be accountable for decisions and supervised work, and record activities | The inspector, not the software, owns the decision |
| RBI code 4.4 | Advice and decisions free from self-interest, prejudice or bias | Watch for bias in risk scoring or training data |
| RBI code 5.8 | Record, retain and disclose evidence, instructions, advice, judgements and decisions | Keep the AI output and your review of it |
| Annex 1.2 | Show how decisions and opinions were reached, for audit | An unexplained AI output is not an auditable reason |
| OSR 3.8 and 3.10 | Reports for site-based and remote inspections; date/time marked, geo-tagged evidence | Remote and AI-processed images need provenance |
Other provisions matter in particular tasks. Confidentiality (RBI code 5.5, RBCA rule 5.6) and commercial data storage (2.12 and 2.20) apply when drawings go into third-party tools; see AI and applicant confidentiality. The OSRs require a risk-based approach (1.1), up-to-date systems and controls (1.2) and accurate records kept for at least 15 years (1.7). RBCA rule 5.9 requires evidence, instructions, advice and decisions to be kept for 15 years from instruction, or longer if insurance requires. Under RBI code Annex 1.10 a potential code breach must be notified within 14 days.
Many building control surveyors are RICS members, so the RICS standard on responsible use of AI in surveying practice applies to them as well. It has been mandatory since 9 March 2026 for members and regulated firms. Where an AI output has a material impact on the service, it expects a written register of AI systems, a risk register reviewed at least quarterly, a documented decision on the reliability of each relied-upon output by a named, appropriately qualified surveyor, and advance written notice to clients that AI will be used. These records fit well with the BSR expectation of auditable decisions, so a firm that meets the RICS standard is better placed under the codes too.
CROSS-UK Report 1554 (16 July 2026) described an AI-written calculation that reached the design stage without competent verification. The panel said approving authorities may wish to ask how far AI contributed to submissions, and that designers should disclose AI-assisted outputs. This is a safety panel view, not a BSR rule, but it signals two things for building control: the documents you assess may have been produced by AI, and your own reliance on AI should be as transparent as you would expect from a designer. The designer's side of a higher-risk submission is covered in AI in Gateway 2 applications for principal designers.
In practice, a building control body can apply the codes to AI with these steps:
See AI plans assessment and AI-written inspection records for task-specific controls.
A breach of the codes is a regulatory matter, but the same facts usually support a negligence claim by a client, owner or purchaser. PI is generally designed to respond to negligence claims, subject to the policy terms, while regulatory fines and the cost of responding to an investigation may or may not be covered depending on the wording. The codes require RBCAs and self-employed RBIs to be suitably insured but set no minimum limit. See does building control PI cover AI errors? and PI insurance for building inspectors and approvers.
Apex arranges PI for building control approvers and inspectors, and a broker can check how your wording treats AI-assisted decisions and any fire safety restrictions.
No. As at October 2026 none of the RBI code, the RBCA professional conduct rules or the operational standards rules mentions artificial intelligence, and the BSR has not published separate AI guidance. Technology, competence, accountability and record duties still apply to AI tools.
RBI code 2.11 and RBCA rule 2.19 deal directly with technology: be trained in it and identify and mitigate its risks. In practice RBI code 4.2 on accountability and Annex 1.2 on auditable decisions are just as important when an AI output informs a decision.
It applies to RICS members and RICS-regulated firms, including those doing building control, and has been mandatory since 9 March 2026. Building control professionals who are not RICS members are not bound by it, but its record-keeping approach is a useful model.
No BSR sanction, disciplinary decision or court ruling involving AI use by a building inspector or RBCA had been published as at October 2026. BSR published its disciplinary panels policy on 21 October 2026.
Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.
Start a PI quote Request a callbackApex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.
