Privacy and Security
SENEX starts from a local-first boundary: personal context should remain within the user-controlled environment unless the user approves a specific use or exchange. The goal is not merely to move storage onto a device, but to make data flows purposeful, limited, visible, and revocable.
Local context handling, user-directed workflows, and governed-sharing foundations are under active development in AIA. These foundations reduce unnecessary central collection, but they are not a formal privacy guarantee or security certification.
Public privacy model
SENEX organizes privacy around four questions:
- Where does information begin? Personal context begins inside the user-controlled environment.
- Why is it being used? Access should be tied to an understandable user purpose.
- What crosses a boundary? Only the minimum information needed for an approved task should be exchanged.
- How does control end? Continuing access should be reviewable and revocable.
This model covers raw content and derived information. A summary, classification, or recommendation can still reveal sensitive facts, so derived results should receive the same purpose and scope review as their sources.
Trust boundaries
- User environment: AIA, local data, device access, and user permissions. Security here also depends on the device, operating system, backups, credentials, and user configuration.
- Approved recipient: another participant or service that receives a bounded result. SENEX cannot make a recipient forget information already disclosed, so approval must precede transfer.
- Coordination network: the future V1-testnet should receive only the information required to coordinate and account for an event, not unrestricted personal context.
- Operators and maintainers: software updates, support processes, and operational access require their own controls and review.
See Privacy Technologies for the capability-level control framework and Threat Model for the risks considered.
Status boundaries
Network privacy behavior is a future test target. V1-testnet will use test-only value and is not a live public blockchain or production privacy service. Its purpose is to expose design and operational failures before any V1 decision.
Advanced collaboration protections remain a research direction. Public docs do not claim that formal privacy properties, secure collective computation, or anonymity guarantees are currently implemented.
Assurance, not slogans
Privacy claims should be tied to evidence: documented data flows, observable permission behavior, repeatable tests, issue handling, and independent review. “Local,” “encrypted,” or “decentralized” are not complete security conclusions on their own.
Independent assessment is required for privacy boundaries, authorization, software updates, network behavior, and any future formal claim. See Compliance for legal and regulatory limits. SENEX does not currently claim certification under a privacy, security, healthcare, or financial standard.
Implementation-specific cryptography, internal interfaces, thresholds, infrastructure choices, and exploitable operating details are intentionally excluded from public documentation.