What is insider threat?

An insider threat is the risk that a person with legitimate access, such as an employee, contractor, administrator or service provider, harms an organization's systems or data with that access, deliberately or by mistake. Most insider incidents come from mistakes rather than malice, and the same access reaches backups.

What kinds of insider threat exist?

Insider threats fall into four groups, defined by intent and by who is actually using the account:

  • Negligence. Mistakes by people doing their jobs: the wrong retention policy attached to a job, a repository removed during a cleanup, a shared administrator password, multi-factor login turned off. The Ponemon Institute 2025 Cost of Insider Risks report found that 55% of insider incidents came from employee negligence or mistakes.
  • Privilege misuse. Access used beyond its purpose without intent to destroy, such as reading or copying data the role does not need, or retrieving files from backups outside an approved request.
  • Rogue administrators. Deliberate harm by someone with high privileges, such as deleting backup chains or shortening retention before leaving the organization.
  • Compromised insiders. An outsider using a legitimate login obtained through credential theft. In the logs, the activity looks identical to the real person's.

Contractors and managed service providers appear in each group: one technician account that manages backups for many clients is an insider at every one of them. For recovery, the motive makes no difference, because a deleted backup chain is gone whether an error or a decision removed it.

Why are insiders hard to detect?

An insider already has access, so there is nothing to break into and perimeter controls do not apply. Backup administrators hold some of the most powerful access in an organization. From the console they can delete restore points, shorten retention or disable jobs, and each of those is also ordinary maintenance. That is why backup tampering by an insider is hard to tell apart from routine work.

How does one account reach both data and backups?

Separation of duties means no single person can both make and approve a critical change, and it depends on having enough people to hold separate roles. Where the same few administrators run production, virtualization and backup, one account can delete both the data and the backups that would restore it, a wide administrative blast radius. An approval step with no one available to approve it does not narrow that radius.

How is insider damage limited?

Storage that refuses deletion until a retention date, as WORM storage does, applies the same rule to every account. With S3 Object Lock in compliance mode, no account, root included, can delete a locked backup early, so a mistake, a deliberate act and a borrowed login all meet the same limit. It does not stop reading or copying data, or misuse of production systems and the backup software itself. Departures remain a high-risk moment, because access often outlasts the role.

Frequently asked questions

Is an insider threat always an employee?

No. Contractors, service providers and outsiders using a borrowed login all appear as insiders in the logs.

Does separation of duties solve the problem?

Only where there are enough people to separate the roles. Where roles cannot be split, storage that refuses early deletion covers the gap.

Does immutability stop insider threats?

It limits what an insider can delete or change. It does not stop reading data or misuse of production systems and the backup software.