Loading...
Sector · Architecture
AI for architects, with design authority kept in the studio
We build AI for architects that speeds up research, drafting and admin, while the architect keeps design authority, regulatory sign-off and professional liability.
In focus →
Live routingHover to pause
Incoming
Checking…
AI handles0
A person decides0
UK architecture practices using AI0%Up from 59% in 2025. RIBA AI survey 2026.
Source: Royal Institute of British Architects, 22 July 2026 ↗Respondents reporting better practice productivity0%
Source: Royal Institute of British Architects, 22 July 2026 ↗Respondents who agree their designs are better because of AI0%
Source: Royal Institute of British Architects, 22 July 2026 ↗Where the line sits
Most practices already use AI, few have decided where it stops
AI in architecture has moved from experiment to everyday practice.
Practices use it for early visuals, specification drafting, bids and admin.
The harder question is where it should stop.
Compliance with building regulations, design changes on live jobs and advice to clients all carry professional and legal weight.
We help practices put AI for architects to work on the volume and keep those calls with a named architect.
What the AI handles
- Planning policy research and precedent summaries
- First drafts of specification clauses and reports
- Bid and proposal writing from your past submissions
- Early concept images for workshops and briefs
- Meeting minutes, action lists and correspondence drafts
What stays with the architect
- Design authority and the final design
- Building regulations and fire safety compliance
- Sign-off on changes during construction
- Advice to clients and contract decisions
- Anything covered by professional indemnity
Where it helps
AI for architects is already part of practice
The RIBA AI Report 2026 found nearly three quarters of surveyed practices use AI on at least some projects.
Respondents report productivity gains, but far fewer say their designs are better for it.
That matches where we see AI for architects earn its place in a practice.
It is strong at research, first drafts, bid writing and admin, and useful for quick concept images.
Bid and tender responses show the pattern well: AI drafts answers from your past submissions, and a named person reviews and approves them before anything is submitted.
Practices that bid for UK public-sector work can also use dedicated AI RFP software, keeping the same rule that a person signs off what goes out.
It is weak at the synthesis, judgement and accountability that make a design good and safe.
Our AI workshops help a studio test these uses on its own projects before committing.
- Planning and policy research with citations
- Specification and report drafting
- Bid writing from your past submissions
- Minutes and correspondence from meetings
85%
AI acts alone
130
65% of 200 cases
Goes to a person
70
35% of 200 cases
Illustrative batch with a typical confidence spread. Your real split comes from a pilot on your own work. Some decisions go to a person whatever the score.
What stays with you
Design authority, compliance and liability stay with the architect
The Architects Code sets standards for competence and the responsible exercise of professional judgement.
AI for architects does not change who is accountable for the work.
Under the building regulations dutyholder regime, designers must be competent and principal designers must assess whether design work complies.
An AI tool cannot hold that competence or carry that duty.
RIBA also notes that questions on professional indemnity, liability and intellectual property remain largely unresolved.
So we design workflows where AI drafts and a named architect checks, edits and signs.
Each workflow keeps a record of what the AI produced and what the architect changed.
AI drafts
Check
Person decides
Learns
01 / 04
Governance
Adoption has run ahead of practice policy
RIBA says governance in practice has not kept pace with adoption, with few practices having formal AI policies.
RIBA plans an AI Overlay to the RIBA Plan of Work and a responsible use policy in 2026.
You do not need to wait for them to set sensible rules in your own studio.
A short policy should cover client confidentiality, which tools staff may use and who checks AI output.
It should also cover how you tell clients where AI has been used.
Clear rules give your team more time for design and client work, with a record of who checked what.
Our AI policy template is a practical starting point.
RIBA says governance in practice has not kept pace with adoption, with few practices having formal AI policies.
How it works
Four steps, one owner
Each system ships with a named person accountable for what it does.
01
01
Audit current use
We find out which AI tools your team already uses, on which projects and with what client data.
02
02
Pick the first workflows
We choose two or three high-volume tasks, such as bids or specifications, and set clear review points.
03
03
Build and write the rules
We set up the tools, connect them to your own material and draft a studio AI policy with you.
04
04
Train the studio
We train architects and assistants to use, check and challenge AI output on live work.
Questions
Asked and answered
Further reading
From the source
Royal Institute of British ArchitectsRIBA Artificial Intelligence Report 2026 ↗The profession’s largest survey of AI use in practice.Royal Institute of British Architects, 22 July 2026AI adoption reaches tipping point across UK architecture ↗Royal Institute of British ArchitectsRIBA’s Adrian Malleson responds to the 2026 AI Report ↗Liability, insurance and governance issues the report leaves open.Architects Registration BoardArchitects Code: Standards of Conduct and Practice ↗GOV.UKDesign and building work: meeting building requirements ↗Dutyholder and competence duties for designers and principal designers.
Keep reading
Related pages
TrainingAI workshops →An AI workshop that ends with a plan, not a demoGovernanceAI policy template →An AI policy template your team will actually followTrainingAI training for business →AI training for business, with people who know when to step inConsultancyAI for small business →An AI consultant for small business, that starts small and stays safe
Next step
Tell us the work. We’ll show you where the line goes.
A 30-minute call. We map one workflow, what the AI could take, and where a person must stay in the loop.