Security Model

Threats

• compromised user device;
• stolen wallet or communication key;
• malicious recipient;
• phishing and impersonation;
• malicious relay;
• compromised storage provider;
• dependency attack;
• insecure browser extension;
• metadata correlation;
• spam and Sybil attacks;
• malicious file;
• group administrator abuse;
• smart-contract vulnerability.

Controls

• end-to-end encryption;
• forward-secrecy-capable protocol design;
• device verification;
• safety-number or key-fingerprint comparison;
• key-rotation alerts;
• signed software releases;
• hardware-wallet support;
• strict content-security policy;
• link and file warnings;
• transparent block and report systems;
• smart-contract audits;
• reproducible builds where possible;
• bug-bounty program;
• minimal telemetry;
• independent security review.

Security warnings

Null cannot prevent:

• a recipient from copying content;
• malware from reading decrypted content on a compromised device;
• users from approving malicious transactions;
• public settlement data from appearing on a public chain;
• voluntary identity disclosure;
• all forms of network-level observation.