flowleak

Flowleak is a term used in information security to describe the unintended leakage of sensitive information through the data flow paths of a software system. It occurs when data representing a confidential source is propagated to outputs, logs, or other components that are not authorized to access it, even if no explicit exfiltration occurs. The concept emphasizes how information can travel through a program via variables, function calls, or integrated services, creating an avenue for leakage beyond conventional access controls.

Flowleaks commonly arise from imperfect taint tracking, misconfigured logging, verbose error reporting, or inadequate data redaction. They can involve direct exposure in API responses, aggregation layers, or dashboards, as well as indirect channels such as timing side channels or resource usage that reveal sensitive state.

Examples include a web API returning a user identifier along with a non-sensitive field in a general error message, an admin interface logging secrets in debug logs, a multi-tenant service combining inputs in a shared analytics pipeline that exposes protected data to unauthorized users, or a timing attack that reveals information based on response latency tied to secret values.

Detection and mitigation for flowleak rely on information-flow analysis tools that track taint from sensitive sources to sinks, often combining static and dynamic analysis. Mitigations include secure-by-default logging with redaction, enforcing least-privilege access, separating data stores, minimizing data exposure, disabling verbose production logging, and implementing explicit data-handling contracts. Ongoing auditing and monitoring help identify emerging leakage paths.

See also: information flow security, taint analysis, data leakage, side-channel attack.