Leak is the unauthorized disclosure or exposure of information to people or systems not meant to receive it. It may be deliberate or accidental and can involve personal data, credentials, API keys, source code, trade secrets, or internal documents. A leak can result from theft and publication, an employee sending data to the wrong recipient, or an exposed cloud storage bucket, database, log, repository, or backup. The term describes the exposure, not necessarily how attackers obtained it; reporting may refer to both confirmed disclosure and suspected exposure.
Security teams should establish what data was accessible, to whom, and for how long, while distinguishing evidence of access from mere exposure. Exposed passwords, tokens, and keys should be revoked or rotated quickly, and affected systems checked for reuse or further access. Personal or regulated data may trigger privacy and reporting obligations, while leaked proprietary material can require legal and threat-intelligence monitoring. Prevention includes least-privilege access, secret scanning, safe sharing controls, encryption where appropriate, and monitoring for misconfigured public resources.