Cyberattacks are becoming increasingly sophisticated, automated, and difficult to detect. Today, attackers rely on sophisticated tactics such as phishing, “living off the land” techniques, legitimate remote tools, or stealthy malware to infiltrate networks undetected and establish a foothold there. For companies, this means that traditional security measures alone are no longer sufficient.
This is exactly where SySS’s Digital Forensics & Incident Response (DFIR) department comes in. By analyzing security incidents and malware, the team investigates attack methods, reconstructs compromises, and identifies indicators and tactics used by attackers. This blog post analyzes a recent piece of malware that the DFIR department has already found three times on customer systems, and shows how modern malware is structured and what insights can be derived from it for defense.
Easy2Convert (hxxps://www[.]easy2convert[.]com/) is a legitimate suite of programs that has been in existence since 2008. The programs are designed to convert various file types, including PCD to TIFF, BMP to JPG, and RAW to JPG.
Attackers often attempt to disguise their malware as legitimate software to evade detection. In the second half of 2025, the Easy2Convert malware (hxxps://www[.]ez2convertapp[.]com/) first appeared. The attackers are thus using the name of an existing legitimate application to distribute their own malware. The first entry in the Wayback Machine (a collection of archived snapshots of websites) dates back to June 2025.
At first glance, it looks like a legitimate website. Unfortunately, what is provided is hardly a legitimate tool. Even though the actual functionality exists and works in part, there are other highly suspicious features in the automatic update function.
The website promises that the downloaded program can convert ODT files to PDF files, merge PDF files, and that any subsequent downloads will take place immediately. This immediately raises several questions:
The first question is the easiest to answer: Many people who work in office jobs and need to convert Word files to PDFs probably don’t know that Word has this built-in capability. When they search for it on Google, various tools appear in the “Sponsored” tab of the search results. (This feature is enabled by default for Google searches.) A program that does everything I want – and even more? Well, bring it on! It’s therefore likely that the attackers’ distribution strategy relies on the lack of knowledge among people who aren’t very familiar with computers.
The other two questions are harder to answer off the cuff. This blog post first addresses the legitimate functions of the three samples found, followed by an analysis of the malicious component. What’s fascinating is that all three instances differ in their legitimate functions. However, the malicious component is exactly the same in all three. The malware’s tactics and methods are classified using the MITRE ATT&CK framework.
All three samples are different: in their functions and, in some cases, in the programming language used.
| Sample | Wrapper Language | Conversion Method | Payload | Payload Language |
| Sample 1 | C# .NET | Upload to an online conversion website | Easy2ConvertUpdater.exe | C# .NET |
| Sample 2 | Rust | Upload to an online conversion website | Easy2ConvertUpdater.exe | C# .NET |
| Sample 3 | Rust | Local Python environment | Easy2ConvertUpdater.exe | C# .NET |
It’s immediately apparent: There are drastic differences between the various implementations. Three different ways to achieve the same result, in two different wrapper languages.
The program is delivered as an installer. This installer then places all required files in the "AppData\Local" directory. These include the program itself, the updater, a unique user ID, various DLL files, and an XML file.
The program actually offers legitimate functions, which will be discussed later. The updater acts as a loader for code that is downloaded later. It is therefore the actual malware that downloads and executes code from the Internet. More on the updater later. The unique user ID is used by the updater. The DLL files provide some of the functionality required by both the updater and the legitimate application. The XML file contains information for a scheduled task. The malware uses this scheduled task to persist and is executed once a day.
What’s interesting about this malware is that the functionality it promises actually exists – at least in part. The conversion from ODT to PDF files works, but the merge function does not. The claim that downloads occur immediately likely applies to one of the three samples. One of the three samples actually uses an online converter and then downloads the converted file again.
The legitimate functions of the samples, in terms of their conversion methods, are described below.
The first version uses .NET to perform the conversion. This feature can even convert other file types (DOC, DOCX, PNG, JPG, PPT, PPTX, XLS, XLSX):
Here you can see that this version can only convert DOCX files after all. It’s noticeable that the user interface (UI) accepts more file types than are supported on the conversion page. It’s unclear what happens to the other file types. The download is then simply saved to the current user’s “Downloads” folder:
BUT: There is no merge function, but there is a download function.
The second version takes the ODT file and uploads it to an online conversion tool via its API. After the file has been converted, it is downloaded to the user’s “Downloads” folder. So two out of three promises are actually fulfilled here. That’s when the problems start, though: Not only is this method problematic from a data privacy perspective, but it also relies on an internet connection and on the website actually being accessible. One could argue that this keeps local dependencies to a minimum and reduces the overall size of the binary. In addition, the antivirus scanner may flag this activity as suspicious and remove the program for that reason alone.
Overall, this is therefore not a good solution for converting ODT files.
The third version includes the functionality to create a Python environment and then use it to convert the ODT file to a PDF file. This uses the Python library “pandoc,” which can perform various conversions from one markup language to another. This is a much more elegant method for converting file formats. However, antivirus scanners may still flag this as suspicious. It is rather unusual for “normal” programs to create and use a Python environment on their own. Once again, the merge functionality is missing here, and this time, the “fast” download is missing as well.
Why does the application have so many different “wrappers” – that is, different legitimate functions across versions – even though the stager is the same for every version? We can only speculate here. It stands to reason that these changes were made so that the installer’s hash would change, thereby preventing it from being filtered out by static detection before installation. The updater has not changed, so it could be detected in the same way every time. Consequently, there is definitely room for improvement here on the part of the attackers.
Users download the stager themselves (MITRE T1204). Presumably, the website purchased ads on Google – for example, when searching for “Word to PDF.” Since the Command and Control (C2) server is no longer accessible at the time of writing this blog post, this assumption unfortunately can no longer be verified. The actual C2 communication takes place via an updater in a scheduled task (MITRE T1053).
The great thing about this Scheduled Task file is that it reveals quite a bit about the malware.
It can be assumed that the author used a template, did not clean it up, and as a result, there are some unnecessary settings or duplicates. Of course, some of the settings may also indicate valid functionality, since the settings mentioned above can also be used in valid updaters.
Now that we know how the updater is invoked, we’ll examine its functionality. Since it’s also written in .NET, dnSpy (a .NET debugger and assembly editor) was used to view the original code. The following function names stand out:
The exact program flow is as follows:
The user ID, which is set during installation, is first JSON-serialized along with the installation timestamp and the current timestamp, and then sent to the "/check-for-updates" endpoint. The timestamps follow the pattern yyyMM-d HH::mmss*. It is unclear why the timestamp looks this way. It is unlikely that this is obfuscation, as this is not the case anywhere else. It can be assumed that this is necessary because the timestamp is used for AES encryption. Another possibility is that dnSpy misinterpreted something here.
If the request fails, the program terminates. If the request succeeds, the same data is sent again to the "/get-update" endpoint. The data returned by the server is encrypted. To decrypt it, the installation time is hashed and used as a key. The "now" timestamp is used as the initialization vector. The decrypted content is then loaded directly into memory.
Finally, the MD5 hash of the message is sent to the "/status-ok" endpoint – along with the user ID – so that the system can verify whether the "update" was successful. If the “/get-update” endpoint does not respond or the request fails for another reason, the error message is sent to the “/status-bad” endpoint along with the user ID.
So what are we dealing with here? An unusual, homemade encryption scheme; code that is loaded directly into RAM from a peculiar URL that has nothing to do with the application’s name. There are many similarities to legitimate updaters, but the evidence pointing to malware outweighs them. Unfortunately, it’s impossible to reconstruct exactly what is loaded into memory, since the C2 server (hiko.lakienti) no longer exists at the time of writing this blog post.
Detection turns out to be relatively straightforward: There is always a Scheduled Task and files in the “AppData/Local/Easy2Convert” directory. Using these files and their hashes, you can relatively quickly determine whether you are actually affected. You can often also observe the execution in the event logs.
Interestingly, in some cases, execution is simply stopped due to a TLS error. Since a Windows update in November 2025, any attempt at execution has failed. It stands to reason that Windows prohibits .NET from using certain TLS versions for web requests. A second update in November 2025 updated Defender. Since then, it has been detecting and removing the Easy2ConvertUpdater.
There is some evidence that this is indeed malware: the peculiar Scheduled Task maneuvers designed for persistence, as well as the URL from which the updates are retrieved – which has nothing to do with the application name and no longer even exists. Furthermore, the question arises as to why a user ID would need to be sent along or why encryption would need to be performed again with timestamps in addition to TLS.
Another indicator is this statement in the Terms of Service on the website:
Here, the attackers attempt to protect themselves from potential legal consequences by including a disclaimer.
The analysis shows how attackers can exploit legitimate software components – and update mechanisms in particular – to conceal communication with compromised systems. Since update services are considered trustworthy in many environments, suspicious network traffic often goes more unnoticed than with traditional malware.
The following insights can be drawn for defense purposes:
Ultimately, the analysis illustrates that modern attackers are increasingly relying on the misuse of existing functions rather than solely injecting their own malware components. For defenders, this means that understanding legitimate system and application processes is just as important as analyzing classic malware.
The following is a list of the Indicators of Compromise – that is, the forensic artifacts found:
Ihr direkter Kontakt zu SySS +49 7071 407856-9107 oder anfrage@syss.de | Sie haben einen Cybersicherheitsvorfall? +49 7071 407856-99
Ihr direkter Kontakt zu SySS +49 7071 407856-9107 oder anfrage@syss.de
Sie haben einen Cybersicherheitsvorfall? +49 7071 407856-99
Direkter Kontakt
+49 7071 407856-9107 oder anfrage@syss.de
Sie haben einen Cybersicherheitsvorfall?