Data Management & Privacy Model

Sietal’s approach to data privacy begins with strict minimization. Only the data that’s absolutely necessary for a given interaction is processed. Before anything is shared, the user’s agent applies a multi-layered filter: removing identifiers, hashing sensitive fields, and evaluating contextual permissions.

All data is encrypted using modern symmetric and asymmetric cryptography (e.g., AES-256, elliptic curve), and tokenized through a data vault system that assigns each data object an access policy and traceable ID. These policies define who can access the data, for how long, and under what smart contract conditions.

Agents log every access attempt and use decision trees to verify if the request aligns with the user’s defined rules. If there’s a violation, say, a third-party dApp attempts to extract profile data without permission, the agent denies access and alerts the user. These logs are verifiable and stored off-chain, with optional hash commitments on-chain for auditability.

Most importantly, Sietal offers retroactive revocation. If you decide that a service should no longer use your data, your agent can issue a signed denial. While already-used data can’t be removed from past computation, the agent ensures that no future usage is possible. This introduces ethical data boundaries rarely seen in decentralized platforms.