Trust and Assurance
Understand how Search Intelligence handles evidence, information and responsibility.
3 min read
Search Intelligence asks organisations to govern material evidence carefully. The same standard must apply to Search Intelligence’s own claims, website, research, client information, personal information, analytical conclusions and publication decisions.
Assurance areas
Privacy and personal information
What may be collected, why it is needed, who may receive it and how rights can be exercised.
Cookies and website technologies
Which technologies are necessary, which are optional and how preferences can be changed.
Data security and confidentiality
Data minimisation, access, suppliers, retention, incidents and AI-assisted processing.
Editorial and evidence integrity
Doctrine, observation, interpretation, modelling, validated evidence, illustrative content and promotion.
Accessibility
Target, assessment status, priority journeys, known limitations and barrier reporting.
Vulnerability disclosure
A controlled route for suspected security vulnerabilities affecting approved SI-controlled systems.
Website terms and conditions
Terms governing use of the Search Intelligence website, public content, resources and intellectual property.
Access to information
PAIA Manual, Information Officer details, records and prescribed request routes.
These pages support transparency. They do not automatically provide certification, a formal assurance opinion, a replacement for a client agreement or proof that every risk has been removed.
The core evidence-handling principles
Data minimisation
SI requests only what is relevant to the agreed decision question. A broader request is not automatically a better one.
Access by need
Access is limited to approved people and suppliers who require it to perform the agreed work.
Human accountability
Technology and AI may support analysis. People remain accountable for interpretation, evidence status, materiality and publication.
Permission-controlled publication
Client names, findings, numerical claims and case evidence are not published without permission or an approved anonymisation basis.
AI-assisted processing
AI may support organisation, comparison, summarisation, drafting and preliminary analysis.
AI may not approve its own output, invent evidence, infer permission, expose confidential information, or present Inferred evidence as Validated. Client-confidential information is not intentionally submitted to public AI tools without an approved basis and agreed controls.
Procurement and due diligence
The SI Procurement and Assurance Due-Diligence Pack identifies the formal disclosures that should be completed before appointment: legal entity, tax and insurance, suppliers, hosting and data location, AI use, IP and evidence ownership, retention, incidents, continuity, conflicts, and termination.
Download the Procurement and Assurance Due-Diligence Pack