Trust & security

BrandKind does not use your content to train AI models, and neither do the third-party AI processing providers we use. Where a provider offers a choice of underlying models, we configure that choice to exclude any option that would use submitted content for training, and disable optional logging or improvement features. Where a provider's no-training protection is set at account level rather than applied by default, we have set that option and keep a record of it. See the privacy policy for detail.

Our approach to security is to describe only what is genuinely in place and to claim no certifications we do not hold. Data is encrypted both while it travels to and from the service and at rest in our database. The safeguards that protect access to your data are:

• Encryption in transit: your connection to BrandKind is protected with HTTPS/TLS, so anything sent between your browser and the service is encrypted on the way.
• Encryption at rest: your data is encrypted on our database provider's infrastructure while it is stored, not only in transit.
• Authenticated, role-based access: you sign in with an email and password, and what you can see or do is governed by your role in the workspace.
• Per-tenant data isolation: each workspace's data is separated from every other at the database layer, so no customer can reach another's content.

BrandKind is multi-tenant: many organisations use the same service, but each has its own separate workspace. That separation is enforced by the database itself, not by application code alone. Every request is scoped to the workspace it belongs to, so one organisation's content can never be returned to another. It is the same mechanism that keeps your drafts private to your own team.

Your account and saved content are held in the United Kingdom and stay there at rest. The text you submit for rewriting is handled by third-party AI and machine-learning processing providers, some of which operate outside the UK, including in the United States, under the international-transfer safeguards set out in our privacy policy. [CONFIRM: one AI processing provider's transfer-safeguard documentation has not yet been supplied to us as a signed copy — do not treat this as confirmed for that provider until we hold it.]

The account emails the service sends you, such as password resets, workspace invitations and service notifications, go through a separate email delivery provider. It receives your email address and the contents of those messages. The text you write and generate is not sent to it: where a notification tells you a piece has finished, the email carries that piece's title and a link to open it in BrandKind, never the content itself.

A full list of our sub-processors, naming each provider, is available on request under a data-processing agreement (DPA). See who we share data with for the categories of provider involved.

You can get your content out of BrandKind at any time. The service is built around producing text you copy straight into your own channels, so nothing is locked in. Export is per piece — you copy a piece out or download it as a document — rather than a single button that packages the whole workspace at once. You can delete individual pieces yourself, and you can ask us to delete your whole workspace, which removes your organisation's data from the service.

Deleting a piece moves it to your Archive, where it can be restored for 30 days before it is deleted permanently. Content you have not deleted is never removed on a timer.

A data-processing agreement is available for customers who need one for their own compliance, and it carries the named list of our sub-processors. Request it from info@roodtsquared.co.uk.

If you believe you have found a security vulnerability in BrandKind, please tell us so we can fix it. Report it to info@roodtsquared.co.uk, with enough detail for us to reproduce the issue. Please give us a reasonable chance to respond before disclosing it publicly.