Privacy Assurance and Evidence
Formal privacy is a property of a precisely defined system under explicit assumptions. It cannot be established by naming a cryptographic technique, publishing an equation, or describing a future architecture. SENEX therefore separates current product controls from research goals and independently validated guarantees.
The current AIA foundation emphasizes local context handling, deliberate exchange, and user control. These are meaningful design constraints, but they do not constitute a mathematical privacy guarantee.
What a formal claim would require
Before SENEX can make a formal privacy claim, it must define:
- Protected information: exactly which user facts, source records, or derived attributes the claim covers.
- Adversary: what the observer can access, control, combine, and retain.
- System boundary: which software, devices, network participants, and operators are inside or outside the claim.
- Assumptions: conditions that must remain true for the claim to hold.
- Composition: how privacy changes when a user repeats actions or combines features.
- Failure behavior: what users and operators can observe and do when a control fails.
Those definitions must match the deployed system. A proof about an abstract component does not automatically cover the surrounding application, logs, metadata, endpoint security, recipient behavior, or operational process.
Evidence levels
SENEX uses the following public distinction:
| Evidence level | Meaning |
|---|---|
| Design intent | A desired property has been documented |
| Implemented control | A control exists in a specific build and scope |
| Internally tested | Repeatable tests support a bounded claim |
| Independently assessed | An external specialist has reviewed the defined scope |
| Formally supported | The deployed mechanism and assumptions support a stated formal property |
Only evidence from the relevant build and deployment should be used. Roadmap text, laboratory prototypes, and testnet observations cannot be promoted into V1 guarantees without a new assessment.
V1-testnet is intended to generate evidence about real integration and operational behavior using test-only value. It is not itself a formal privacy guarantee and will not automatically transfer its state or assurance to a future V1 network.
Research boundary
Privacy-enhancing cooperation is a Research direction. SENEX may investigate mechanisms that reduce disclosure during multi-party work, but public documentation intentionally omits constructions, parameters, thresholds, and evaluation details that could expose proprietary design or weaken security.
No current SENEX documentation should be interpreted as proof of anonymity, non-inference, unlinkability, confidential collective computation, or suitability for regulated data. Any such claim requires a defined scope, implementation evidence, independent assessment, and clear limitations.
This evidence-first approach is deliberately conservative: it gives users a way to distinguish what is intended, what is implemented, and what has actually been validated.