What researchers saw
A valid login gave the attacker a way to run the next stage
FortiGuard Labs says it uncovered an active campaign in early October targeting internet-facing Microsoft SQL Server instances. The malware scanned for exposed servers, tried supplied user names and passwords, and used a successful login to execute malicious T-SQL. The commands downloaded or assembled additional malware on the compromised Windows host.
The researchers traced follow-on components that created scheduled tasks, collected host information, and supported cryptomining and remote control. Their report describes a particular campaign and its samples; it does not establish that every exposed SQL Server is infected or that this is a new SQL Server software vulnerability.
Treat this as an access and execution chain
Defender workflow
Follow the login into the host
- 1
Find exposed SQL services. Identify servers reachable on SQL Server ports from untrusted networks. Review whether remote administration and SQL authentication are required, and limit access to approved clients.
- 2
Review authentication and query activity. Look for bursts of failed logins followed by a successful session, then unusual administrative T-SQL or command-execution behavior. Correlate database logs with network and endpoint telemetry.
- 3
Check the Windows endpoint. Investigate unexpected downloads, processes launched from temporary locations, scheduled tasks masquerading as update tasks, and outbound contact to unfamiliar hosts. Use FortiGuard's published indicators as pivots, not proof on their own.
Defender note
Do not stop the investigation at a failed-login alert
Credential guessing is noisy, but the important turn is the first accepted login and what the database session could do afterward. If that session ran attacker-controlled T-SQL, scope the Windows host as well as the database account; rotate exposed credentials and check for persistence before returning the service to normal.