top of page

Responsible AI in construction

AI that supports professional judgement

Responsible AI means we introduce AI to support the people accountable for the work, never to replace them. Every engagement starts by understanding your data, agreeing who can access what and deciding where people must review the output.

Our approach to responsible AI

01   Understand the data

We map what information a workflow uses, where it is held and how sensitive it is before any tool is configured.

02   Agree access

Access boundaries are agreed with you and limited to what each workflow needs.

03   Keep people in the loop

Outputs used in commercial or contractual decisions are reviewed by a qualified person, with the review step built into the workflow.

04   Measure honestly

We agree how success and errors will be measured, and we report both.

Data and security in our engagements

Confidentiality

We work under a non-disclosure agreement as standard.

Your data, your control

Client information is used only for the agreed engagement and stays within the systems and accounts you approve.

Data protection

We work in line with UK GDPR and the UAE Personal Data Protection Law.

Access and deletion

Access is removed and working copies are deleted at the end of an engagement, as agreed in writing.

Tool and provider choices

We are not tied to one provider. Model, hosting and retention settings are reviewed and confirmed with you during scoping.

Security questions

Security and governance questions can be sent to hello@vantusai.co.uk.

Have a governance question?

We are happy to talk through how a workflow would handle your data before any work begins.

Frequently asked questions

Which AI models and providers do you use?

We are not tied to one provider. Choices are made for each engagement based on your requirements and confirmed in writing.

Where is our data hosted?

Hosting and data residency are agreed during scoping and depend on the tools selected and your own policies.

Do you keep our data?

Retention is agreed in each engagement. By default, working copies are deleted when the engagement ends.

Is our data used to train AI models?

Tool settings and provider terms are reviewed during scoping so this can be confirmed before any of your data is used.

Do you hold security certifications?

Not at present. We explain our working practices openly and are happy to complete your supplier questionnaire.

Start with a conversation

Tell us about a workflow that takes too long or carries too much risk. We will talk it through and suggest a sensible first step.

bottom of page