What is LLM data residency?
It is the question of where information associated with an LLM service may be processed or stored. The answer can depend on more than the model provider: other services and infrastructure may be involved.
TRUST BOUNDARY INTELLIGENCE
Know where LLM data goes.
Know how it is handled.
Trust Boundary builds evidenced intelligence on large language model data residency and handling. We make the data pathway easier to examine, so organisations can ask better questions about the AI they rely on.
Built for scrutiny.
A clear distinction between what is said, what is evidenced, and what remains to decide.
LLM DATA PATHWAY
ORGANISATIONAI useWhat is in use?
SERVICELLM providerWhich service?
HANDLINGData pathwayWhere and how?
A CHANGING DATA PATHWAY
AI is not one place.
It is a chain of handling.
An LLM interaction can involve more than the model name on the screen.
LLM data residency concerns where prompts, outputs and related information may be processed or stored. Data handling concerns what happens to that information along the way. Services, infrastructure and operating practices can all matter. Trust Boundary brings relevant evidence together so people can examine those questions.
What we mean by evidenced intelligenceWhere is information processed?
Where may it be stored or retained?
Who else may handle it along the way?
FROM CLAIMS TO CONTEXT
Evidence should
make the picture
more legible.
Trust Boundary intelligence is designed to bring third-party evidence to questions of LLM data residency and handling. It helps show what the evidence relates to, where it has limits, and what still needs human judgment.
Trace the question
Make the data-residency or handling question explicit.
Relate evidence
Keep supporting information distinct from an organisation’s own statement.
Keep limits visible
Evidence informs a decision. It does not make one for you.
ILLUSTRATIVE RECORD STRUCTURE
QUESTIONS + EVIDENCEQuestions worth evidencing
Residency
Where is data processed, stored, or otherwise located?
Handling
What happens to prompts, outputs, and related information?
Evidence
What third-party information supports or leaves the question open?
Example structure only. No provider finding or supplier assessment is represented here.
Claim ≠ evidenceIntelligence ≠ certificationDecision remains with the buyer
DataPathby Trust Boundary
Put the intelligence
to work in a tender.
DataPath is a separate product proposition that applies Trust Boundary intelligence to a buyer’s tender context.
DATAPATH · A RELATED APPLICATION
Part of the Risk Assessment
A supplier may use AI anywhere in its business, whether or not AI is part of the service offered. DataPath takes the supplier’s attestation about AI use and its data-residency claim, then uses Trust Boundary intelligence to give the buyer relevant third-party evidence.
SUPPLIERAttestationAI use + residency claim
TRUST BOUNDARYIntelligenceEvidence relevant to claim
BUYERConsiderationDecision stays with buyer
ABOUT TRUST BOUNDARY
Why Trust Boundary
exists.
Trust Boundary was founded to address a practical gap: organisations increasingly rely on large language models without always having a clear view of where their information may travel or how it is handled. Data-residency claims should be examined alongside relevant third-party evidence, not simply repeated.
By making the questions, evidence and its limits visible, Trust Boundary helps people consider what is known, what remains uncertain and what they must decide. The aim is better-informed judgment, not a badge or a promise of certainty.
Adam McInnes B.Sc.
Founder and Chief Data Scientist
Adam brings 30+ years of experience in procurement and scaling technology businesses. His current work spans SaaS, AI, procurement and cybersecurity.
Adam on LinkedInDATAPATH PILOT INTEREST
Take part.
Shape the pilot.
We are inviting Australian government procurement buyers to register their interest in a DataPath pilot. Share your contact details and we can get in touch about the opportunity to take part.
Registering interest is not a commitment.
Participation and scope are discussed before pilot work begins.
Interested in taking part?
Register your interest in the DataPath pilot with just your contact details. The form creates no account.
Express pilot interestThe questions behind the evidence.
What does Trust Boundary provide?
Trust Boundary develops intelligence that relates third-party evidence to specific questions about LLM data residency and handling, while making the evidence limits visible.
How does DataPath use that intelligence?
DataPath is a separate tender application. It connects a supplier’s AI-use attestation and data-residency claim with relevant Trust Boundary intelligence for an Australian government buyer to consider.
TRUST BOUNDARY