Clear controls. Honest limits.

Security and governance are part of how we design and deliver AI work. We consider the information involved, who should have access, what the system is allowed to do and where human review is needed.

The detailed arrangements depend on the platforms, features and integrations selected. Before an implementation, we agree the relevant access, data handling, responsibilities and support arrangements with the client.

Connected services may involve additional providers and their own terms. We identify these as part of the design rather than assuming that one platform's commitments cover the entire workflow.

AI outputs can be incorrect or incomplete. Material outputs need appropriate review before they are relied on. We do not promise zero risk, guaranteed compliance or universal data-location commitments.