Aipedia: Building control approvers

AI in building control: rules, risks and PI

Aipedia, by the Apex broking team · Last reviewed 2026-10-30

Building control professionals can use AI tools, but the Building Safety Regulator expects registered inspectors and approvers to understand and manage the risks of any technology they rely on. Because RBIs and RBCAs remain accountable for every decision, AI errors become their professional liability.

Part of: AI in building control approvers

In short

The Building Safety Regulator (BSR) has published no AI-specific guidance for building control as at October 2026. Its codes still apply to AI: RBI code standard 2.11 requires training in, and risk management of, technology; RBCA professional conduct rule 2.19 requires firms to identify and mitigate technology risks; and the operational standards rules cover remote inspections (3.8) and require time-stamped, geo-tagged evidence (3.10). Inspectors stay accountable for decisions (4.2) and must keep auditable records. RBCAs and self-employed RBIs must be suitably insured, but no minimum PI limit is set.

How building control uses AI

Building control work is document heavy and evidence driven, which makes it an obvious target for AI tools. Registered building inspectors (RBIs), registered building control approvers (RBCAs, formerly approved inspectors) and local authority building control teams in England are using or trialling tools for:

Plans assessment: rule-checking software that reads drawings and specifications against the Approved Documents and flags possible gaps.
Summarising: generative AI that condenses long fire strategies, structural reports and higher-risk building applications.
Site inspection: image recognition on site photos, video calls for remote inspections and dictation apps that turn voice notes into reports.
Planning and admin: tools that score project risk and propose inspection frequency, draft letters and notices, and answer applicant questions through chatbots.

The regulator itself has changed. The Building Safety Regulator, which registers RBIs and RBCAs and is the building control authority for higher-risk buildings (HRBs), became a standalone arm's-length body sponsored by MHCLG on 27 January 2026, having previously sat within HSE. On 9 July 2026 it announced a more proportionate and targeted approach to HRB assessments. Neither announcement addresses AI.

The rules that apply

As at October 2026 the BSR had published no guidance, rule or statement on AI in building control, and none of the codes mentions artificial intelligence. The technology and accountability duties in the codes apply to AI tools in the same way as any other software.

Source	Applies to	Status	Duty relevant to AI
RBI code of conduct	All RBIs	Binding (updated 22 May 2026)	2.11 be trained in and manage risks of technology; 3.1 competence; 4.2 accountability; 5.8 records
RBCA professional conduct rules	RBCAs	Binding (updated 22 May 2026)	2.19 identify and mitigate technology risks; 5.6 confidentiality; 5.9 records for 15 years
Operational standards rules	Building control bodies	Binding (updated 22 May 2026)	3.8 site-based and remote inspections; 3.10 time-stamped, geo-tagged evidence; 1.7 records for 15 years
RICS AI standard	RICS members and regulated firms	Mandatory from 9 March 2026	AI register, risk register, written reliability decisions, client notice
CROSS-UK Report 1554	Industry (safety panel view)	Not binding (16 July 2026)	Approving authorities may wish to ask how far AI contributed to submissions

Other standards matter as much as the technology clauses. Under the RBI code an inspector must only carry out restricted activities they are registered and competent for (3.1), is accountable for decisions and supervised work (4.2), and must give advice free from bias (4.4). Annex 1.2 requires records to show how decisions and opinions were reached so they can be audited. A decision that rests on an AI output nobody can explain sits awkwardly with all of these. The BSR codes and AI guide maps each provision.

The main risks

The main risks for building control are:

False assurance at plan check: a rule-checker that passes a drawing can encourage a lighter human review, especially on fire safety (Part B) and structure.
What the camera did not see: remote inspections and AI image tools rely on footage the builder chose to send.
Inaccurate records: AI-drafted inspection notes that are wrong may undermine a defence many years later, when claims can arise under the Building Safety Act's extended limitation periods.
Certificates on incomplete evidence: OSR 4.3 requires every contravention to be resolved before a completion or final certificate is issued.
Confidentiality and data: uploading drawings and fire strategies to third-party tools engages RBI code 5.5, RBCA rule 5.6, security-sensitive HRB information and UK GDPR.
Competence and registration: AI-written portfolio evidence that overstates competence can lead to work outside registration scope, which Annex 4.1 makes an offence.

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, so the first outcomes are worth watching.

The PI position

RBCAs must hold adequate insurance, give BSR the certificate on request and only take on work they are suitably insured for. Self-employed RBIs must maintain suitable insurance, and terms of engagement must state the insurance held. The codes do not name PI or set a minimum limit, policy terms or run-off, so what is "suitable" is a judgement for the firm and its insurer. Our guide to PI insurance for building inspectors and approvers covers the basics.

A claim that an inspector negligently missed a breach is usually treated as a professional negligence claim whether or not AI was involved, subject to the policy terms. The harder questions are exclusions and disclosure. Construction PI policies may restrict cover for fire safety and cladding, which can bite on exactly the defects AI tools are marketed to find, and some policies now include AI-specific clauses. Insurers may ask how AI is used in plans assessment and inspection at renewal. See does building control PI cover AI errors? and the wider question of whether PI covers AI mistakes.

Apex arranges PI for building control approvers and self-employed inspectors. A broker reviews your current cover, including fire safety restrictions and how the wording treats AI-assisted plan checks and inspections.

Aipedia guides for building control

Each guide covers one task or issue in detail.

BSR codes and AI: what inspectors must do: how the RBI code, RBCA rules and operational standards apply to AI tools.
AI plans assessment against the regulations: rule-checkers reading drawings, and why the decision stays with the RBI.
AI-assisted remote and video inspections: what a camera cannot show and the evidence rules in OSR 3.10.
AI defect detection in inspection photos: missed fire stopping, false negatives and over-reliance on flags.
AI-written inspection reports and records: 15-year records and why accuracy matters years later.
AI and completion or final certificates: compiling completion evidence without certifying on gaps.
AI review of Gateway 2 applications: triaging large higher-risk building submissions within competence.
AI risk-based inspection planning: algorithmic risk scores, key stages and auditable decisions.
AI-drafted contravention notices and referrals: wrong regulation references and invalid notices.
AI in fire service and statutory consultations: summarising fire strategies without losing the engineering detail.
AI and regularisation of unauthorised work: owner-supplied material and hidden elements.
AI in RBI competence portfolios and CPD: evidence that must reflect your own competence.
AI chatbots for building control applicants: advice, identity checks and terms of engagement.
Does building control PI cover AI errors?: what "suitably insured" means and how policies may respond.
AI, applicant data and confidentiality: drawings, fire strategies and security-sensitive HRB information.

Sources

Frequently asked

Has the Building Safety Regulator issued AI guidance?

Not as at October 2026. The BSR has published no AI-specific guidance, rule or statement for building control, and the codes do not mention artificial intelligence. The technology duties in RBI code 2.11 and RBCA rule 2.19, and the record and accountability duties, apply to AI tools in the same way as other software.

Can an RBI rely on an AI plan-checking tool?

An RBI can use it as an aid, but remains accountable for decisions under RBI code 4.2 and must only carry out restricted activities they are registered and competent for (3.1). Records must show how a decision was reached (Annex 1.2), so the inspector's own reasoning needs to be on file.

Is there a minimum PI limit for building control approvers?

No. RBCAs must hold adequate insurance and work only where suitably insured, and self-employed RBIs must maintain suitable insurance, but the BSR codes do not name PI or set a minimum limit, policy terms or run-off requirement.

Should I tell my insurer I use AI?

Usually yes, if it is material to the risk. Under the duty of fair presentation you should disclose how AI is used in plans assessment, inspection and reporting, and check whether your policy contains AI or fire safety clauses that could affect cover.

Check your PI covers how you use AI

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 callback

Apex 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.