Detecting Obfuscated Command-lines with a Massive Language Mannequin


Within the safety business, there’s a fixed, indisputable fact that practitioners should cope with: criminals are working time beyond regulation to always change the menace panorama to their benefit. Their methods are many, they usually exit of their strategy to keep away from detection and obfuscate their actions. Actually, one ingredient of obfuscation – command-line obfuscation – is the method of deliberately disguising command-lines, which hinders automated detection and seeks to cover the true intention of the adversary’s scripts.

Varieties of Obfuscation

There are a couple of instruments publicly out there on GitHub that give us a glimpse of what methods are utilized by adversaries. Considered one of such instruments is Invoke-Obfuscation, a PowerShell script that goals to assist defenders simulate obfuscated payloads. After analyzing a few of the examples in Invoke-Obfuscation, we recognized completely different ranges of the method:

Every of the colours within the picture represents a special method, and whereas there are numerous kinds of obfuscation, they’re not altering the general performance of the command. Within the easiest type, Gentle obfuscation adjustments the case of the letters on the command line; and Medium generates a sequence of concatenated strings with added characters “`” and “^” that are usually ignored by the command line. Along with the earlier methods, it’s doable to reorder the arguments on the command-line as seen on the Heavy instance, by utilizing the {} syntax specify the order of execution. Lastly, the Extremely stage of obfuscation makes use of Base64 encoded instructions, and by utilizing Base8*8 can keep away from a big quantity EDR detections.

Within the wild, that is what an un-obfuscated command-line would appear like:

One of many easiest, and least noticeable methods an adversary might use, is altering the case of the letters on the command-line, which is what the beforehand talked about ‘Gentle’ method demonstrated:

The insertion of characters which are ignored by the command-line such because the ` (tick image) or ^ (caret image), which was beforehand talked about within the ‘Medium’ method, would appear like this within the wild:

In our examples, the command silently installs software program from the web site The method used on this case is particularly stealthy, since it’s utilizing software program that’s benign by itself and already pre-installed on any pc operating the Home windows working system.

Don’t Ignore the Warning Indicators, Examine Obfuscated Components Rapidly

The presence of obfuscation methods on the command-line usually serves as a robust indication of suspicious (virtually at all times malicious) exercise. Whereas in some situation’s obfuscation might have a sound use-case, comparable to utilizing credentials on the command-line (though this can be a very unhealthy thought), menace actors use these methods to cover their malicious intent.  The Gamarue and Raspberry Robin malware campaigns generally used this method to keep away from detection by conventional EDR merchandise. For this reason it’s important to detect obfuscation methods as rapidly as doable and act on them.

Utilizing Massive Language Fashions (LLMs) to detect obfuscation

We created an obfuscation detector utilizing giant language fashions as the answer to the always evolving state of obfuscation methods. These fashions encompass two distinct components: the tokenizer and the language mannequin.

The tokenizer augments the command strains and transforms them right into a low-dimensional illustration with out dropping details about the underlying obfuscation method. In different phrases, the aim of the tokenizer is to separate the sentence or command-line into smaller items which are normalized, and the LLM can perceive.

The tokens into which the command-line is separated are primarily a statistical illustration of widespread mixtures of characters. Subsequently, the widespread mixtures of letters get a “longer” token and the much less widespread ones are represented as separate characters.

It is usually vital to maintain the context of what tokens are generally seen collectively, within the English language these are phrases and the syllables they’re constructed from. This idea is represented by “##” on the planet of pure language processing (NLP), which suggests if a syllable or token is a continuation of a phrase we prepend “##”. One of the simplest ways to show that is to take a look at two examples; Considered one of an English sentence that the widespread tokenizer gained’t have an issue with, and the second with a malicious command line.

For the reason that command-line has a special construction than pure language it’s crucial to coach a customized tokenizer mannequin for our use-case. Moreover, this practice tokenizer goes to be considerably higher statistical illustration of the command-line and goes to be splitting the enter into for much longer (extra widespread) tokens.

For the second a part of the detection mannequin – the language mannequin – the Electra mannequin was chosen. This mannequin is tiny when in comparison with different generally used language fashions (~87% much less trainable parameters in comparison with BERT),  however remains to be in a position to study the command line construction and detect beforehand unseen obfuscation methods. The pre-training of the Electra mannequin is carried out on a number of benign command-line samples taken from telemetry, after which tokenized. Throughout this part, the mannequin learns the relationships between the tokens and their “regular” mixtures of tokens and their occurrences.

The following step for this mannequin is to study to distinguish between obfuscated and un-obfuscated samples, which is named the fine-tuning part. Throughout this part we give the mannequin true constructive samples that have been collected internally. Nevertheless, there weren’t sufficient samples noticed within the wild, so we additionally created an artificial obfuscated dataset from benign command-line samples. Throughout the fine-tuning part, we give the Electra mannequin each malicious and benign samples. By exhibiting completely different samples, the mannequin learns the underlying method and notes that sure binaries have the next chance of being obfuscated than others.

The ensuing mannequin achieves spectacular outcomes having 99% precision and recall.

As we regarded via the outcomes of our LLM-based obfuscation detector, we discovered a couple of new methods identified malware comparable to Raspberry Robin or Gamarue used. Raspberry Robin leveraged a closely obfuscated command-line utilizing wt.exe, that may solely be discovered on the Home windows 11 working system. Alternatively, Gamarue leveraged a brand new methodology of encoding utilizing unprintable characters. This was a uncommon method, not generally seen in experiences or uncooked telemetries.

Raspberry Robin:


The Electra mannequin has helped us detect anticipated types of obfuscation, in addition to these new methods utilized by the Gamarue, Raspberry Robin, and different malware households. Together with the prevailing safety occasions from the Cisco XDR portfolio, the script will increase its detection constancy.


There are numerous methods on the market which are utilized by adversaries to cover their intent and it’s only a matter of time earlier than we come across one thing new. LLMs present new prospects to detect obfuscation methods that generalize nicely and enhance the accuracy of our detections within the XDR portfolio. Let’s keep vigilant and hold our networks protected utilizing the Cisco XDR portfolio.

We’d love to listen to what you suppose. Ask a Query, Remark Beneath, and Keep Related with Cisco Safety on social!

Cisco Safety Social Channels