FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Apex Insurance Brokers
Speak to a brokerGet a quote →

Aipedia: Building control approvers

AI chatbots for building control applicants

Chatbots can answer routine applicant questions and start an instruction, but RBI code 5.1 to 5.3 requires identity checks, a record of services and agreed written terms before work starts. Wrong advice to a householder can become a complaint or a claim.

In short

Building control bodies use chatbots and AI portals to tell applicants whether work needs approval, explain the process and take instructions. The RBI code requires confirmation of the applicant's identity, a record of the services requested and an explanation of the services (5.1), a statement of registration limits (5.2), and written agreement to terms of engagement before any work (5.3). Annex 2 lists what the terms must contain, including insurance held. A chatbot that tells a householder approval is not needed, when it is, can cause loss.

How chatbots are used in building control

Many enquiries to building control are simple: does this extension need approval, what does it cost, what happens next. Chatbots and AI-assisted portals can answer these at any time, collect project details, generate quotes and prepare an engagement. Some can also answer technical questions from guidance and suggest whether work is exempt or needs an application.

That last use is where the risk lies. Whether work needs approval can depend on details a chatbot does not ask about, and householders are likely to act on the answer.

The duties that apply

Several existing duties apply to this work:

See the RBI code of conduct and RBCA conduct rules. If a chatbot made a decision about an individual solely by automated means, the safeguards in section 80 of the Data (Use and Access) Act 2025 could apply.

What can go wrong

A chatbot may tell a householder that a garage conversion or a new opening in a wall does not need approval, missing that it affects fire escape or structure. It may describe the service in terms that do not match the engagement letter, or move an applicant to payment before written terms are agreed. Answers drawn from out-of-date guidance can be confidently wrong.

Illustrative: a householder asks an RBCA's website chatbot whether removing an internal wall needs building control. The chatbot answers that internal alterations usually do not. The wall is loadbearing, the work goes ahead without approval, and cracking follows. The householder complains that the RBCA's own tool told them approval was not needed and asks who will pay for remedial work and regularisation.

How PI usually responds

Negligent advice given through a chatbot is likely to be treated like other advice from the firm, subject to the policy terms. Some insurers may ask specifically about client-facing AI tools, and some wordings may have AI clauses. Free general information given before an engagement may raise questions about whether the policy treats it as professional services. See AI chatbot insurance and does building control PI cover AI errors?.

Practical controls

You can reduce the risk with these steps:

See also AI and regularisation and the AI in building control hub. Apex arranges PI for building control approvers, and a broker can check how your policy treats advice given through chatbots. See PI insurance for building inspectors and approvers.

Sources

Frequently asked

Can a chatbot take building control instructions?

It can collect details and start the process, but RBI code 5.1 to 5.3 requires identity confirmation, a record of the services requested and written agreement to the Annex 2 terms of engagement before any work begins.

Must terms of engagement mention insurance?

Yes. Annex 2 of the RBI code requires the terms to include details of the relevant insurance held by the self-employed RBI, if applicable, and by the building control body. RBCA terms must state the relevant insurance or professional liability cover held.

Is a building control body liable for chatbot advice?

It may be, if the advice was negligent and someone relied on it. Whether PI responds depends on the policy terms and how it treats advice given outside a formal engagement. Limiting the chatbot to process information reduces the risk.

Should a chatbot tell applicants whether work is exempt?

It is safer not to. Whether work needs approval can turn on structural, fire or drainage details that a chatbot may not ask about. Route those questions to an inspector, and keep the chatbot to process, fees and general information with a clear statement that it is not a decision.

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.