Read-write access is a permission level that lets a process or user both open files and change them. In cyber security, it is more dangerous than read-only access because it expands the impact of a compromise from simple disclosure to active tampering. An attacker who can only read data may steal documents or keys; an attacker who can write can also alter scripts, replace configuration files, plant persistence, or corrupt evidence.
In sandboxed apps and virtual machines, read-write access to a host filesystem mount is a major trust-boundary risk. If an untrusted workload can reach shared host files, the sandbox may no longer contain the threat. Defenders reduce this risk by using least privilege, preferring read-only mounts, separating secrets from shared paths, and monitoring for unexpected file changes.