When Ransomware Targets the Backups
Description: Reporting by BleepingComputer describes ransomware incidents in which attackers targeted backup and recovery data. The cases illustrate a core resilience question: whether recovery copies remain usable when attackers can reach the systems and identities that protect them.
Ransomware can disrupt an organization even when it has backups. The decisive issue is whether those copies remain intact, accessible to trusted responders, and capable of restoring business operations in a reasonable time. Attackers understand this dependency. Encrypting production data is one form of coercion; destroying recovery points or disrupting restoration infrastructure can remove the victim’s alternative to paying.
The practical lesson is not simply to buy more storage or make another copy. Organizations need to understand the paths that connect production systems, backup platforms, administrative identities, and disaster-recovery locations. If one compromised account or network route can reach all of them, apparent redundancy may offer little protection.
What Attacks on Backup Infrastructure Look Like
BleepingComputer’s account describes several examples of attackers targeting recovery capabilities. The incidents differ, but they share a pattern: threat actors seek access to backup data or the systems that enable restoration, often by abusing credentials or weaknesses in the surrounding environment.
In February 2024, the ALPHV/BlackCat group attacked Change Healthcare after entering through a remote-access portal that reportedly lacked multifactor authentication. The report says the backups were not isolated or robust enough to restore operations quickly. UnitedHealth paid a reported $22 million ransom and the data was still not returned; total recovery costs were estimated at $1.6 billion. These details underline that payment does not guarantee data recovery, and that operational restoration can remain costly even after a ransom is paid.
The report also recounts BlackMatter activity against NEW Cooperative and Crystal Valley in 2021. Attackers using compromised administrator credentials reportedly found backup data stores and appliances on the network and wiped or reformatted them before encrypting other systems. When backup infrastructure is reachable from the same environment and governed by the same privileged identities, a production compromise can become a recovery compromise as well.
A further example in the report concerns Gunra ransomware. A joint CISA and FBI advisory in August 2026 documented a case in which attackers deleted backup and archived data at both a primary data center and a disaster-recovery site. According to the report, a single set of stolen credentials could reach both locations. Physical distance did not provide meaningful resilience where both sites trusted the same credentials.
These cases should not be treated as proof that every ransomware incident follows the same playbook. They do show why defenders should evaluate the full recovery chain, including identity, network access, management consoles, storage settings, and the people authorized to make destructive changes.
Ransomware Backup Security Starts With Mapping Access
A useful review begins with specific questions. Which identities can create, modify, expire, or delete backup copies? Can those identities sign in from ordinary workstations or production servers? Are backup consoles exposed to the same identity provider and network paths as the systems they protect? Can an attacker who compromises a domain or cloud administrator use that access to alter retention settings or disable recovery services?
The answers reveal whether separation is substantive or merely organizational. A backup appliance in another building is not necessarily isolated if it accepts the same credentials and management traffic. Likewise, a cloud copy may remain vulnerable if a compromised administrator can change access policies or delete the account that stores it.
Document the dependencies: administrative accounts, service accounts, management interfaces, network connections, encryption keys, and recovery procedures. Identify which operations are possible from each identity, and make high-impact actions—such as shortening retention or deleting protected copies—require stronger controls than routine backup monitoring. Review this map when systems, vendors, or staff responsibilities change.
Why Backup Defenses Fall Behind
Backup environments are sometimes treated as appliances rather than critical software systems. That can leave them with weaker monitoring, outdated components, or less rigorous access review than production systems. The source report notes that backup software may be patched late and that an attack on an airline using Akira ransomware exploited a vulnerability for which a patch had been available for more than a year.
A separate weakness is shared access. If the same administrator account controls both production and backup infrastructure, theft or misuse of that identity can expose both. Broad privileges also make mistakes and malicious changes more consequential. Organizations should inventory privileged and service accounts, remove unnecessary permissions, protect administrative sign-ins with multifactor authentication, and use dedicated accounts for backup administration.
Monitoring gaps compound the risk. Backup systems should generate and retain logs for sign-ins, configuration changes, retention-policy edits, deletion attempts, and unusual backup or restore activity. Alerts need an owner and a response path. If alerts from a backup console are ignored or never reach incident responders, the organization may discover the attack only when it tries to recover.
Resource constraints matter too. Smaller IT teams and managed service providers may understand the exposure but lack the budget, personnel, or operational time to address every control at once. A prioritized plan can start with the most damaging single points of failure: identities that can delete all copies, unpatched internet-accessible management interfaces, and backups that have never been restored in a test.
Protecting the Way Back
Treat recovery infrastructure as critical infrastructure, not a secondary copy of production. Separate backup administration from everyday IT administration, restrict management access to approved and monitored paths, and avoid shared credentials where one stolen secret could reach every recovery location. Multifactor authentication and role-based permissions add friction, but they must be paired with careful account lifecycle management and review of emergency access.
Use immutable or otherwise protected copies where appropriate. Write-once retention and object-lock capabilities can prevent ordinary administrative actions from altering or deleting copies during their protected period. Configure those controls deliberately: confirm who can change retention, how long copies remain protected, and what process is required for exceptional changes. Immutability is not a substitute for securing accounts, monitoring activity, or testing restoration, but it can limit the damage a compromised account can do.
Keep backup platforms and their dependencies current. Include backup servers, agents, storage interfaces, and management software in the normal vulnerability and patch-management process. Prioritize exposed or high-privilege systems, and track exceptions with a remediation date rather than allowing backup components to become invisible technical debt.
Preserve more than one recovery path. Separate copies by administrative boundary and, where feasible, by location or technology so that a single compromised identity or control plane cannot erase every option. Offline or logically isolated copies can help, provided procedures to access them are documented and practiced. The objective is independent recovery capability, not a particular number of copies on paper.
Test Restores, Not Just Backups
A successful backup job is evidence that data was written; it does not prove that systems can be restored within business requirements. Regularly restore representative files, applications, and complete services into a controlled environment. Verify data integrity, application dependencies, access to required encryption keys, and the time needed to bring the service back.
Include scenarios in which ordinary production credentials or management systems are unavailable. Confirm that responders know how to access protected copies without relying on the compromised identity or infrastructure. Record test results, failures, owners, and follow-up actions. A restore test is valuable only when the organization fixes the problems it uncovers and repeats the test after significant changes.
Incident plans should also specify who can authorize recovery actions, how to preserve evidence, how to communicate with affected teams, and how to prioritize services. Rehearsing these decisions reduces confusion when people are under pressure. The aim is not to promise that every incident will be painless; it is to establish a credible route from detection to safe restoration.
The Cost of Losing Recovery Options
The source article cites IBM’s 2025 Cost of a Data Breach Report, which put the average cost of a ransomware incident at $5.08 million, and reports that IBM’s 2026 research found 41% of ransomware incidents involved threats to damage the victim’s brand reputation. These figures are attributed to the reporting and should be interpreted in their stated context; they do not predict the cost of any one organization’s event.
Operational disruption, recovery work, customer confidence, and regulatory obligations can all shape the impact. When attackers destroy usable recovery copies, responders may face a longer and more uncertain restoration effort. Paying a ransom does not guarantee that data or services will be restored, as the Change Healthcare example in the report illustrates. Resilience therefore depends on controls and rehearsals put in place before an incident, not on assumptions made during one.
A Practical Recovery Readiness Review
Start by listing critical services and the recovery point and recovery time each can tolerate. Map every backup copy and the identities, networks, consoles, and keys needed to restore it. Identify any account or control plane capable of deleting or disabling all copies, then reduce its reach and strengthen its protection.
Next, confirm that protected copies cannot be casually rewritten, that backup software is patched, and that activity is monitored by the security team. Run a restore exercise for a priority service, including a scenario where production credentials are compromised. Turn the results into assigned remediation work and a schedule for retesting.
The central test is straightforward: if an attacker gained control of a powerful production account today, could that account also erase every recovery path? If the answer is yes or unknown, the organization has an actionable resilience gap. Ransomware backup security is ultimately about preserving an independent, tested way back—not merely counting copies.