# BoldMove: FortiOS Expertise and Mandiant’s China-Based Assessment

- **Canonical URL:** https://www.probackend.com/ai-nation-state-cyber-operations/custom-boldmove-backdoor-targets-fortios-devices-in-state-sponsored-campaign
- **Section:** News & Analysis
- **Published:** 2026-11-19
- **Summary:** Mandiant says BoldMove shows deep FortiOS knowledge and assesses the attacker may be China-based. Review the evidence and practical defenses.

---

## What Mandiant reported about BoldMove

Mandiant researchers said BoldMove’s FortiOS-specific features indicate substantial knowledge of how Fortinet devices work. Dark Reading reported that researchers connected the malware to exploitation of a Fortinet FortiOS SSL-VPN zero-day and assessed that the attacker appeared to be based in China. That wording matters: this is a qualified assessment, not a public identification of an operator or proof of state direction. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

The reporting describes a customized Linux version intended for FortiOS as well as a Windows version. The reported targets included a European government entity and an African managed service provider. Those details establish the context Mandiant described, but do not by themselves establish the complete victim set, the full duration of the campaign, or the identity of the people responsible. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

## Why BoldMove suggests deep FortiOS knowledge

Mandiant’s central technical observation, as quoted by Dark Reading, was that the Linux variant included features tailored to Fortinet devices. The researchers said implementation of those features showed an in-depth understanding of device operation; some functions appeared rewritten to run on lower-powered devices. This is more specific than saying that the malware was merely compiled for Linux: it points to adaptation for the environment in which the backdoor was meant to operate. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

The distinction is important for defenders. A general-purpose implant may offer familiar functions such as downloading files or opening a reverse shell. A device-tailored implant can also account for the target platform’s particular features and constraints. Mandiant’s reported assessment was that BoldMove’s customized elements reflect knowledge of FortiOS and Fortinet devices. The reporting does not establish that every component was novel, nor that malware sophistication alone identifies its author.

Dark Reading reported that the Linux sample first attempted to reach a hardcoded command-and-control server. If the connection succeeded, it gathered information about the system and relayed it to that server; instructions could then provide the operator remote control of the affected device. The account describes a capability and an observed malware design, not evidence that every targeted device was successfully compromised or controlled. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

## Exploitation and reported targeting

Researchers linked BoldMove to CVE-2022-42475, a vulnerability disclosed by Fortinet in its FortiOS SSL-VPN technology. Dark Reading reported that the flaw allowed unauthenticated arbitrary code execution on affected systems and that Fortinet had said it was aware of at least one in-the-wild exploitation incident. Mandiant’s available telemetry, as described in the report, suggested exploitation associated with BoldMove was occurring as early as October 2022. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

The malware was discovered in a public repository in December, according to the reporting. That discovery and the researchers’ linkage to the vulnerability help explain how the investigation developed; they should not be mistaken for a complete chronology of initial access or every step of an intrusion. The article reports a European government entity and an African managed service provider among the targets, while describing the broader interest as organizations connected with government work. It does not provide a comprehensive victim list. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

A related Windows flavor was also identified. Mandiant had not detected in-the-wild exploitation associated with that version at the time described by Dark Reading. The sample was reported as 32-bit and not designed to run on a Fortinet device. Its existence is relevant to understanding the malware family, but should not be treated as evidence that the Windows sample played a role in the reported FortiOS compromises. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

## What the China-based assessment does—and does not—say

Dark Reading characterized the attacker as appearing to be China-based, attributing that assessment to Mandiant. This language expresses an analytic judgment with uncertainty. It is not equivalent to naming an individual, establishing physical location beyond doubt, or attributing the activity to a specific government. The observed targeting and technical characteristics may inform an assessment, but readers should not turn that assessment into a stronger claim than the source makes. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

The story places BoldMove in the context of espionage activity and wider targeting of internet-facing network security technologies. That context helps explain why firewalls and related infrastructure can be attractive targets: they sit at a network boundary and may provide an attacker access to a strategically useful vantage point. It does not mean that all Fortinet customers were targeted, or that this particular campaign affected every organization with the product. Keep the reported evidence, the researchers’ interpretation, and any broader geopolitical conclusion separate. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

## Defensive practices for FortiOS operators

Organizations that operate FortiOS devices should treat edge appliances as critical assets. Maintain an accurate inventory of FortiGate and other relevant systems, including versions and exposure, and ensure that vulnerability response follows Fortinet’s security advisories and applicable vendor guidance. Apply relevant fixes through established change-control procedures, prioritizing exposed systems and documenting verification. These are general defensive practices; the Dark Reading account does not claim that any one measure would certainly have prevented the reported activity.

Review administrative access to appliances. Limit management interfaces to authorized networks and personnel, remove accounts and access paths that are no longer needed, and use strong authentication and least privilege where supported. Confirm that configuration changes are authorized and that device administration is not broadly reachable from the public internet. These steps reduce unnecessary exposure, although they do not substitute for addressing a vulnerable service.

Preserve and review available device, authentication, and network telemetry. Investigate unexpected administrative activity, unexplained configuration changes, unusual outbound connections, or other deviations from a known-good baseline. Because the reported Linux malware attempted to contact a hardcoded command-and-control server and collect system information, defenders can use those behaviors as investigative context—not as a claim that every suspicious connection is BoldMove. Validate findings against reliable vendor and incident-response information before drawing conclusions. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

If an appliance may be compromised, follow the organization’s incident-response process: preserve evidence, involve appropriate security and network teams, assess the device and connected environment, and use trusted guidance to contain and recover. Avoid assuming that a routine reboot or a single configuration change establishes that an intrusion has been eradicated. The source describes remote-control capability, so assessment should consider both the appliance and the access it provides to surrounding networks, while remaining proportionate to evidence.

## A measured takeaway

BoldMove’s significance in the reporting rests on the combination of a FortiOS-related exploitation context, a Linux implant adapted for Fortinet devices, and Mandiant’s assessment that the attacker appeared China-based. The most concrete lesson is that network appliances warrant disciplined vulnerability management, access control, logging, and incident response. The attribution remains qualified, the reported victim examples are not a complete census, and the malware’s capabilities should not be confused with proof of impact in every environment. [Dark Reading](https://www.darkreading.com/threat-intelligence/china-based-attacker-crafted-custom-malware-for-fortinet-zero-day)

![mandiant reported about boldmove](https://assets.probackend.com/ee62ed49-4145-442e-972d-7babc9cd8d85?token=ast_7b9ee8975a4b187969d396286e3f9717c42f2331)
