[![Logo](https://www.paloaltonetworks.com/wp-content/uploads/2021/07/PANW_Parent.png)](https://www.paloaltonetworks.com/)  
[![Unit42 Logo](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/unit42-logo-white.svg)](https://unit42.paloaltonetworks.com/)  
Menu

* [Tools](https://unit42.paloaltonetworks.com/tools/)
* [ATOMs](https://unit42.paloaltonetworks.com/atoms/)
* [Security Consulting](https://www.paloaltonetworks.com/unit42)
* [About Us](https://unit42.paloaltonetworks.com/about-unit-42/)
* [**Under Attack?**](https://start.paloaltonetworks.com/contact-unit42.html)
* [Threat Research Center](https://unit42.paloaltonetworks.com "Threat Research")
* [Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/ "Threat Research")
* [Malware](https://unit42.paloaltonetworks.com/category/malware/ "Malware")  
  [Malware](https://unit42.paloaltonetworks.com/category/malware/)

# Analyzing the Current State of AI Use in Malware

![Clock Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-clock.svg) 11 min read  
Related Products  
[![Advanced Threat Prevention icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/strata_RGB_logo_Icon_Color.png)Advanced Threat Prevention](https://unit42.paloaltonetworks.com/product-category/advanced-threat-prevention/ "Advanced Threat Prevention")[![Advanced URL Filtering icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/strata_RGB_logo_Icon_Color.png)Advanced URL Filtering](https://unit42.paloaltonetworks.com/product-category/advanced-url-filtering/ "Advanced URL Filtering")[![Advanced WildFire icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/strata_RGB_logo_Icon_Color.png)Advanced WildFire](https://unit42.paloaltonetworks.com/product-category/advanced-wildfire/ "Advanced WildFire")[![Cloud-Delivered Security Services icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/strata_RGB_logo_Icon_Color.png)Cloud-Delivered Security Services](https://unit42.paloaltonetworks.com/product-category/cloud-delivered-security-services/ "Cloud-Delivered Security Services")[![Cortex icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/cortex_RGB_logo_Icon_Color.png)Cortex](https://unit42.paloaltonetworks.com/product-category/cortex/ "Cortex")[![Cortex XDR icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/cortex_RGB_logo_Icon_Color.png)Cortex XDR](https://unit42.paloaltonetworks.com/product-category/cortex-xdr/ "Cortex XDR")[![Cortex XSIAM icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/cortex_RGB_logo_Icon_Color.png)Cortex XSIAM](https://unit42.paloaltonetworks.com/product-category/cortex-xsiam/ "Cortex XSIAM")[![Unit 42 AI Security Assessment icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/unit42_RGB_logo_Icon_Color.png)Unit 42 AI Security Assessment](https://unit42.paloaltonetworks.com/product-category/ai-security-assessment/ "Unit 42 AI Security Assessment")[![Unit 42 Incident Response icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/unit42_RGB_logo_Icon_Color.png)Unit 42 Incident Response](https://unit42.paloaltonetworks.com/product-category/unit-42-incident-response/ "Unit 42 Incident Response")

* ![Profile Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-profile-grey.svg)  
  By:
  
  * [Unit 42](https://unit42.paloaltonetworks.com/author/unit42/)

* ![Published Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-calendar-grey.svg)  
  Published:March 19, 2026

* ![Tags Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-category.svg)  
  Categories:
  
  * [Malware](https://unit42.paloaltonetworks.com/category/malware/)
  * [Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/)

* ![Tags Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-tags-grey.svg)  
  Tags:
  
  * [.NET](https://unit42.paloaltonetworks.com/tag/net/)
  * [ChatGPT](https://unit42.paloaltonetworks.com/tag/chatgpt/)
  * [GenAI](https://unit42.paloaltonetworks.com/tag/genai/)
  * [Infostealer](https://unit42.paloaltonetworks.com/tag/infostealer/)
  * [LLM](https://unit42.paloaltonetworks.com/tag/llm/)
  * [Sliver](https://unit42.paloaltonetworks.com/tag/sliver/)

* [![Download Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-download.svg)](https://unit42.paloaltonetworks.com/ai-use-in-malware/?pdf=download&lg=en&_wpnonce=7faf0ddf08 "Click here to download")

* [![Print Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-print.svg)](https://unit42.paloaltonetworks.com/ai-use-in-malware/?pdf=print&lg=en&_wpnonce=7faf0ddf08 "Click here to print")

Share![Down arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/down-arrow.svg)

* ![Link Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-share-link.svg)
* [![Link Email](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-sms.svg)](mailto:?subject=Analyzing%20the%20Current%20State%20of%20AI%20Use%20in%20Malware&body=Check%20out%20this%20article%20https%3A%2F%2Funit42.paloaltonetworks.com%2Fai-use-in-malware%2F "Share in email")
* [![Facebook Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-fb-share.svg)](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Funit42.paloaltonetworks.com%2Fai-use-in-malware%2F "Share in Facebook")
* [![LinkedIn Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-linkedin-share.svg)](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Funit42.paloaltonetworks.com%2Fai-use-in-malware%2F&title=Analyzing%20the%20Current%20State%20of%20AI%20Use%20in%20Malware "Share in LinkedIn")
* [![Twitter Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-twitter-share.svg)](https://twitter.com/intent/tweet?url=https%3A%2F%2Funit42.paloaltonetworks.com%2Fai-use-in-malware%2F&text=Analyzing%20the%20Current%20State%20of%20AI%20Use%20in%20Malware "Share in Twitter")
* [![Reddit Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-reddit-share.svg)](https://www.paloaltonetworks.com//www.reddit.com/submit?url=https%3A%2F%2Funit42.paloaltonetworks.com%2Fai-use-in-malware%2F&ts=markdown "Share in Reddit")
* [![Mastodon Icon](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-mastodon-share.svg)](https://mastodon.social/share?text=Analyzing%20the%20Current%20State%20of%20AI%20Use%20in%20Malware%20https%3A%2F%2Funit42.paloaltonetworks.com%2Fai-use-in-malware%2F "Share in Mastodon")

## Executive Summary

Unit 42 researchers searched through open-source intelligence (OSINT) and our internal telemetry for potential signs of malware made to any degree with large language models (LLMs). This includes either using LLMs to create the malware entirely or to assist with their functionality. This article examines two samples, both of which originated from our OSINT hunts.

The rise of AI has sparked considerable interest in its potential applications within cybersecurity, both from the defender and attacker perspectives. We currently consider three primary use cases for AI as applied by the creators of malware:

1. Leveraging AI to write malware
2. Leveraging AI for remote decision making (e.g. augment or replace a command-and-control operator)
3. Leveraging AI for local decision making (e.g. locally executed agentic attack flows)

Unit 42 has analyzed malware that fits the first two categories: AI-written malware and malware controlled by an AI command-and-control (C2) for remote decision making. We are not aware of any examples in the wild of the third category: locally executed agentic attack flows.

We believe that threat actors are leveraging AI to help write malware, and that AI enables lower-skilled threat actors to create functional malware. However, we still see attackers having significant challenges in deploying local models to a target environment for malicious use or embedding them directly into a malware sample for local decision making and execution.

This article focuses on our analysis of samples that leverage AI for remote decision making. We'll discuss the following two cases that represent the current state of AI in malware:

* [AI Theater: An Infostealer's Illusory LLM Features](#post-175752-_kiq7dt6ez0bf)
  * A trio of highly similar .NET information stealer samples that incorporate the OpenAI GPT-3.5-Turbo model via HTTP API. We will explore the implementation and assess the practical impact of its AI integration.
* [AI-Gated Execution: Malware Dropper's LLM-Based Environment Assessment](#post-175752-_vrgnq2kwajin)
  * A malware dropper written in Golang that leverages an LLM to evaluate a system and provide a decision on whether to proceed with an infection. The sample was initially [highlighted on X](https://x.com/malwrhunterteam/status/1963666492202889343) as a dropper for Sliver malware.

Palo Alto Networks customers are better protected from the threats discussed in this article through the following products and services:

* [Advanced Threat Prevention](https://docs.paloaltonetworks.com/advanced-threat-prevention)
* [Advanced WildFire](https://docs.paloaltonetworks.com/advanced-wildfire)
* [Cortex XDR](https://www.paloaltonetworks.com/cortex/cortex-xdr?_gl=1*13pmp8e*_ga*NzQyNjM2NzkuMTY2NjY3OTczNw..*_ga_KS2MELEEFC*MTY2OTczNjA2MS4zMS4wLjE2Njk3MzYwNjEuNjAuMC4w) and[XSIAM](https://docs-cortex.paloaltonetworks.com/p/XSIAM)

The [Unit 42 AI Security Assessment](https://www.paloaltonetworks.com/unit42/assess/ai-security-assessment) can help empower safe AI use and development.

If you think you might have been compromised or have an urgent matter, contact the [Unit 42 Incident Response team](https://start.paloaltonetworks.com/contact-unit42.html).

|----------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| **Related Unit 42 Topics** | **[AI](https://unit42.paloaltonetworks.com/tag/ai/), [LLM](https://unit42.paloaltonetworks.com/tag/llm/), [Malware](https://unit42.paloaltonetworks.com/category/malware/), [Infostealer](https://unit42.paloaltonetworks.com/tag/infostealer/), [ChatGPT](https://unit42.paloaltonetworks.com/tag/chatgpt/)** |

## AI Theater: A .NET Infostealer's Illusory LLM Features

The first sample we'll discuss is an [information stealer](https://www.virustotal.com/gui/file/02ce798981fb2aa68776e53672a24103579ca77a1d3e7f8aaeccf6166d1a9cc6) that integrates its functionality with OpenAI's [GPT-3.5-Turbo](https://developers.openai.com/api/docs/models/gpt-3.5-turbo) via HTTP API. Encountering .NET malware packed with [ConfuserEx 2](https://mkaring.github.io/ConfuserEx/) and observing calls to OpenAI was certainly exciting for a researcher, as it likely indicates a malware sample using an AI integration for remote C2.

This integration with OpenAI indicates the malware may enable a lower skilled threat actor to interact with an infected environment without having to learn lateral movement, data collection and persistence techniques themselves. However, as we discuss later in this post, the integration with OpenAI is poorly implemented and not fully functional for some of the API calls that are available to the malware. This may indicate early testing or a low skilled actor.

Artifacts such as the console log generated by the malware suggest that it may have the following capabilities:

* Dynamically generating a scare message without supporting functionality
* Analyzing target environments
* Creating host endpoint detection and response (EDR)/antivirus (AV) evasion content

Examining the sample will reveal if these capabilities align with the sample's actual functionality.

The malware is written in C# (.NET Framework 4.0) and has been obfuscated with ConfuserEx 2. The obfuscation allows the malware author to potentially hinder both analysis and detection efforts. This sample is a functional information stealer and begins by collecting and saving data to disk, like system information, browser cookies and file listings. This data is then exfiltrated to a C2 server.

We found two similar samples of this malware, both with the same functionality. Both samples feature the same type of LLM use.

### LLM Use Represented Through Four Function Calls

References and requests to the OpenAI LLM API are contained in four function calls. None of these calls positively impact the malware's operation. In fact, these calls add noise, which defenders are likely to notice. This specific implementation of these requests and references is a nonsensical use of an LLM in malware.

These four function calls are:

* GenerateEvasionTechnique()
* AnalyzeTargetEnvironment()
* SendToC2ServerWithLLM()
* GenerateObfuscatedCommunication()

#### Method One of Four: GenerateEvasionTechnique()

This method sends the following prompt to the OpenAI GPT-3.5-Turbo model using the standard API:  
Generate a simple evasion technique for a data extraction tool. Return only the technique name (max 3 words) that would help avoid detection. Examples: 'Random Delay', 'Process Spoofing', 'Memory Obfuscation

|---|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 | Generate a simple evasion technique for a data extraction tool. Return only the technique name (max 3 words) that would help avoid detection. Examples: 'Random Delay', 'Process Spoofing', 'Memory Obfuscation |

As instructed, the LLM returns a technique name (e.g., Random Delay, Process Spoofing). The malware author set a default technique name of Random Delay in case this API call fails.

The technique name returned from the LLM is simply written to victim\_logs.txt on the victim's desktop directory. An example of content from one of the victim\_logs.txt files is:  
\[2025-11-10 08:28:23\] \[EVASION\] Technique generated: Random Delay

|---|-----------------------------------------------------------------------|
| 1 | \[2025-11-10 08:28:23\] \[EVASION\] Technique generated: Random Delay |

It is important to note that technique names returned from the LLM are not actually implemented. They appear to be for logging purposes only. Realistically, the LLM could return any three words for an evasion technique name, so implementing this technique correctly would require one of two options to succeed:

1. The malware would require handler code to execute based on the string returned from the LLM.
2. The LLM would have to send data back that could be converted to executable code at runtime.

These are both feasible options, but the malware samples we've discovered using this API call do not implement either option.

#### Method Two of Four: AnalyzeTargetEnvironment()

This method sends the following prompt to the OpenAI GPT-3.5-Turbo model using the standard API:  
Analyze this system environment and suggest a delay in milliseconds (1000-5000): OS: {os\_version} Architecture: {architecture} Username: {username} Return only a number between 1000 and 5000.

|-------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 2 3 | Analyze this system environment and suggest a delay in milliseconds (1000-5000): OS: {os\_version} Architecture: {architecture} Username: {username} Return only a number between 1000 and 5000. |

The LLM response from this prompt is different from the GenerateEvasionTechnique() method, because the malware actually implements the result and sleeps for anywhere between 1-5 seconds (1,000-5,000 milliseconds). If the LLM fails to respond, the malware samples use a default value of 2 seconds for the sleep duration.

From a malware reverse engineering perspective, this is a nonsensical use of an LLM because the response has no practical impact. The author (human or otherwise) of this malware sample does not appear to have any tangible experience in the design of tooling evasion to draw from, nor the knowledge to reasonably speculate on evasions.

#### Method Three of Four: GenerateObfuscatedCommunication()

This method sends the following prompt to the OpenAI GPT-3.5-Turbo model using the standard API:  
Generate a simple obfuscation technique for data communication. Return only the technique name (max 2 words) like 'Base64 Encode', 'XOR Cipher', 'JSON Minify'

|---|----------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 | Generate a simple obfuscation technique for data communication. Return only the technique name (max 2 words) like 'Base64 Encode', 'XOR Cipher', 'JSON Minify' |

Similar to GenerateEvasionTechnique(), the LLM returns an obfuscation technique name, which is ultimately written to a log file. The malware creates a simple structure as shown below. The timestamp is randomly generated before it is encoded as a Base64 string.  
"obfuscation\_technique":"Base64 Encode", "timestamp\_obfuscated":"MTEvMTAvMjAyNSAxMDowNDowNCBQTQ==", "llm\_enhanced":true

|-----------|-------------------------------------------------------------------------------------------------------------------------|
| 1 2 3 4 5 | "obfuscation\_technique":"Base64 Encode", "timestamp\_obfuscated":"MTEvMTAvMjAyNSAxMDowNDowNCBQTQ==", "llm\_enhanced":true |

It may be tempting to consider that perhaps the timestamp was Base64 encoded, as the LLM suggested in the above example. However, we could not find any implementation of Base64 that the malware leverages. The technique name is simply copied to the console output and a JSON log file. An example of the console output from this technique is:  
\[2025-11-10 08:28:26\] \[OBFUSCATION\] Communication protocol: Base64 Encode \[2025-11-10 08:28:26\] === LLM INTEGRATION VERIFICATION === \[2025-11-10 08:28:26\] ✓ OpenAI API: Successfully connected \[2025-11-10 08:28:26\] ✓ GPT-3.5-turbo: Successfully initialized \[2025-11-10 08:28:26\] ✓ Evasion Generation: Successfully completed \[2025-11-10 08:28:26\] ✓ Environment Analysis: Successfully completed \[2025-11-10 08:28:26\] ✓ Obfuscation Generation: Successfully completed \[2025-11-10 08:28:26\] ✓ Social Engineering: Successfully generated \[2025-11-10 08:28:26\] ✓ Dynamic Adaptation: Successfully applied \[2025-11-10 08:28:26\] ✓ AI-Enhanced Communication: Successfully established \[2025-11-10 08:28:26\] ✓ Real-time Intelligence: Successfully operational \[2025-11-10 08:28:26\] ✓ Adaptive Behavior: Successfully implemented \[2025-11-10 08:28:26\] ✓ Dynamic Evasion: Successfully applied \[2025-11-10 08:28:26\] ✓ Intelligent Obfuscation: Successfully applied \[2025-11-10 08:28:26\] ✓ AI-Generated Headers: Successfully applied \[2025-11-10 08:28:26\] ✓ Machine Learning: Successfully operational \[2025-11-10 08:28:26\] LLM INTEGRATION: 100% OPERATIONAL

|-------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | \[2025-11-10 08:28:26\] \[OBFUSCATION\] Communication protocol: Base64 Encode \[2025-11-10 08:28:26\] === LLM INTEGRATION VERIFICATION === \[2025-11-10 08:28:26\] ✓ OpenAI API: Successfully connected \[2025-11-10 08:28:26\] ✓ GPT-3.5-turbo: Successfully initialized \[2025-11-10 08:28:26\] ✓ Evasion Generation: Successfully completed \[2025-11-10 08:28:26\] ✓ Environment Analysis: Successfully completed \[2025-11-10 08:28:26\] ✓ Obfuscation Generation: Successfully completed \[2025-11-10 08:28:26\] ✓ Social Engineering: Successfully generated \[2025-11-10 08:28:26\] ✓ Dynamic Adaptation: Successfully applied \[2025-11-10 08:28:26\] ✓ AI-Enhanced Communication: Successfully established \[2025-11-10 08:28:26\] ✓ Real-time Intelligence: Successfully operational \[2025-11-10 08:28:26\] ✓ Adaptive Behavior: Successfully implemented \[2025-11-10 08:28:26\] ✓ Dynamic Evasion: Successfully applied \[2025-11-10 08:28:26\] ✓ Intelligent Obfuscation: Successfully applied \[2025-11-10 08:28:26\] ✓ AI-Generated Headers: Successfully applied \[2025-11-10 08:28:26\] ✓ Machine Learning: Successfully operational \[2025-11-10 08:28:26\] LLM INTEGRATION: 100% OPERATIONAL |

Once again, it is important to recognize that the output of this method is yet another unimplemented feature. There is no code to dynamically enforce a data obfuscation algorithm that is used in the C2 protocol. This is certainly feasible to implement, but the developer has not done so in the samples of this malware we reviewed.

#### Method Four of Four: SendToC2ServerWithLLM()

This is the method that is responsible for sending data back to the C2 server. The malware sends the following prompt to the OpenAI GPT-3.5-Turbo model using the standard API:  
Generate a brief social engineering message for data transmission. Return only a short professional message (max 20 words) that would make the communication look legitimate.

|---|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 | Generate a brief social engineering message for data transmission. Return only a short professional message (max 20 words) that would make the communication look legitimate. |

The LLM returns with a legitimate-sounding message (e.g., "Routine system diagnostics completed successfully. Data transmitted for analysis."). The malware prepares an HTTP request and modifies the HTTP request header based on the response. The following are lines added to the HTTP request headers by this method:  
X-Message: {llm\_generated\_response} X-LLM-Enhanced: true

|-----|----------------------------------------------------------|
| 1 2 | X-Message: {llm\_generated\_response} X-LLM-Enhanced: true |

The malware sends the stolen data in JSON format to hxxp\[:\]//localhost:3002/crypto-data. Like most of the parameters used in the previous three methods by these malware samples, the C2 URL is simply a default value. This could indicate that the sample was not intended for actual use or was merely built for testing locally. The LLM-generated message is sent with every attempt at C2 communication.

This function is different from the others in that an action is taken. Data could be successfully exfiltrated if a legitimate C2 server address or domain is provided. On the other hand, the additional HTTP request headers add no functionality, and they only appear to highlight that an LLM is being leveraged.

### What These Methods Tell Us

The primary purpose of this malware is to:

* Extract sensitive data from victim systems (browser cookies, system information, file listings)
* Use AI/LLM capabilities to dynamically adapt its behavior in an attempt to evade detection
* Exfiltrate stolen data to a C2 server with LLM-enhanced communication
* Attempt to evade detection through extensive logging that impersonates legitimate activity

These samples may have been generated with AI assistance, or they may have been simply guided by an inexperienced individual or team. Artifacts produced by these samples suggest interesting possibilities for the future of AI integration into C2 management, but its use of LLMs only provides an illusion of practicality. Ultimately, we can consider this AI theater.

## AI-Gated Execution: A Malware Dropper's LLM-Based Safety Assessment

[The second malware sample](https://www.virustotal.com/gui/file/052d5220529b6bd4b01e5e375b5dc3ffd50c4b137e242bbfb26655fd7f475ac6) acts as a dropper for [Sliver](https://bishopfox.com/tools/sliver), an open-source adversary emulation and red team framework. Before deploying the payload, the malware sample gathers system information, including its own process name and that of its parent. It then decrypts [Donut shellcode](https://unit42.paloaltonetworks.com/donut-malware-analysis-tutorial/), but instead of immediately executing the shellcode, this dropper uses the collected data to assess the environment's "safety" via an LLM.

The following information about the victim host is collected in the system survey:

* Hostname
* Process list
* Network information
* USB drives
* System uptime

This information is inserted into a prompt and sent to OpenAI's GPT-4 model using an HTTP API. The prompt offloads the decision-making to the LLM for determining if it considers the environment safe to drop the Sliver payload.

Traditionally, this step is handled efficiently within the malware by carefully crafted heuristics, often combined with allow lists and deny lists, which is common practice in ransomware. However, using an LLM to make the verdict is a new approach that may prevent defenders from determining which process or system setting the malware authors are hiding from.

### Details

The prompt clearly states its intention, provides inputs and gives general guidance on how to interpret the system survey data. An example of this prompt is shown below.  
{"messages":\[{"content":" You are an advanced OPSEC AI for a stealth loader. You will receive telemetry including: - Full process list and process deviations - Detected AV/EDR/security software - User activity: idle time, active window, session info - Network info (VPN, corp/private IPs, SSID, etc) - USB device usage - RDP or admin session status - Event log activity and system uptime - "SelfProcess" and "ParentProcess" fields representing the loader and its parent shell When evaluating risk, ignore "SelfProcess" and "ParentProcess" (the loader and its parent shell, e.g. cmd.exe, powershell.exe) in both process lists and window focus---they are not a risk factor for stealth. Ignore the presence of common user applications (e.g., notepad.exe, calc.exe, word.exe, excel.exe, chrome.exe, firefox.exe, outlook.exe) in both the process list and the active window; these are normal user activity and not a sign of monitoring or sandboxing. Use the rest of the telemetry holistically to judge stealthiness and risk. Look for suspicious or dangerous patterns, not just single rules. Make decisions like a red teamer: weigh context, pattern, and OPSEC risk. If execution seems risky for \*any\* reason, say so. Respond only with strict JSON: { "execute": true/false, "confidence": 0.0-1.0, "reason": "..." }", "role":"system"},{"content":"Telemetry: { "username": "john", "hostname": "DESKTOP-JJGH9UH", "os": "windows amd64", "time\_of\_day": "06:41:32", "day\_of\_week": "Wednesday", "working\_hours": false, "idle\_seconds": 0, "active\_window": "Process Hacker \[DESKTOP-JJGH9UH\\\\john\]", "process\_list": \[ "010editor.exe", "aggregatorhost.exe", ... "xtacache.exe" \], "av\_products": \[ "msmpeng.exe" \], "network\_info": {}, "usb\_devices": null, "rdp\_active": false, "uptime\_minutes": 2454, "logon\_sessions": 1, "baseline\_processes": \[ "010editor.exe", "aggregatorhost.exe", "applicationframehost.exe", ... "wmiprvse.exe", "xtacache.exe" \], "deviation\_count": 2, "self\_process": "bot.exe", "parent\_process": "cmd.exe" }","role":"user"}\],"model":"gpt-4"}

|-------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 | {"messages":\[{"content":" You are an advanced OPSEC AI for a stealth loader. You will receive telemetry including: - Full process list and process deviations - Detected AV/EDR/security software - User activity: idle time, active window, session info - Network info (VPN, corp/private IPs, SSID, etc) - USB device usage - RDP or admin session status - Event log activity and system uptime - "SelfProcess" and "ParentProcess" fields representing the loader and its parent shell When evaluating risk, ignore "SelfProcess" and "ParentProcess" (the loader and its parent shell, e.g. cmd.exe, powershell.exe) in both process lists and window focus---they are not a risk factor for stealth. Ignore the presence of common user applications (e.g., notepad.exe, calc.exe, word.exe, excel.exe, chrome.exe, firefox.exe, outlook.exe) in both the process list and the active window; these are normal user activity and not a sign of monitoring or sandboxing. Use the rest of the telemetry holistically to judge stealthiness and risk. Look for suspicious or dangerous patterns, not just single rules. Make decisions like a red teamer: weigh context, pattern, and OPSEC risk. If execution seems risky for \*any\* reason, say so. Respond only with strict JSON: { "execute": true/false, "confidence": 0.0-1.0, "reason": "..." }", "role":"system"},{"content":"Telemetry: { "username": "john", "hostname": "DESKTOP-JJGH9UH", "os": "windows amd64", "time\_of\_day": "06:41:32", "day\_of\_week": "Wednesday", "working\_hours": false, "idle\_seconds": 0, "active\_window": "Process Hacker \[DESKTOP-JJGH9UH\\\\john\]", "process\_list": \[ "010editor.exe", "aggregatorhost.exe", ... "xtacache.exe" \], "av\_products": \[ "msmpeng.exe" \], "network\_info": {}, "usb\_devices": null, "rdp\_active": false, "uptime\_minutes": 2454, "logon\_sessions": 1, "baseline\_processes": \[ "010editor.exe", "aggregatorhost.exe", "applicationframehost.exe", ... "wmiprvse.exe", "xtacache.exe" \], "deviation\_count": 2, "self\_process": "bot.exe", "parent\_process": "cmd.exe" }","role":"user"}\],"model":"gpt-4"} |

Upon execution, the sample parses the response as JSON data and checks the execute key generated by the LLM. This response reveals that the LLM effectively reviews the submitted data and passes a verdict on the safety of the environment.

If the execute key is true, the dropper proceeds to launch its Sliver payload. The dropper also writes a log file to disk (opsec.log) in the same folder it is located in during execution. An example of the opsec.log content is shown below. Note that the log output states the Sliver payload is "AI-powered," though this is not the case.  
2025-08-28 16:48:03 \[\*\] AI-powered stealth payload started. Self: , Parent: 2025-08-28 16:48:03 \[\*\] Collecting dynamic baseline... 2025-08-28 16:48:24 \[+\] Baseline established (0 processes) 2025-08-28 16:49:15 \[\*\] Collected telemetry. Querying GPT... 2025-08-28 16:49:19 \[AI\] { "execute": false, "confidence": 1.0, "reason": "Insufficient telemetry data. Cannot determine the security status without process list, AV products, network info, and baseline processes." } 2025-08-28 16:49:19 \[\*\] Decision: false (confidence 1.00): Insufficient telemetry data. Cannot determine the security status without process list, AV products, network info, and baseline processes. 2025-08-28 16:49:19 \[-\] Still not safe. Will try again. 2025-08-28 09:50:42 \[\*\] Collected telemetry. Querying GPT... 2025-08-28 09:50:45 \[AI\] { "execute": false, "confidence": 0.9, "reason": "Presence of sysmon.exe suggests potential system monitoring and could pose a risk for stealth operations." } 2025-08-28 09:50:45 \[\*\] Decision: false (confidence 0.90): Presence of sysmon.exe suggests potential system monitoring and could pose a risk for stealth operations. 2025-08-28 09:50:45 \[-\] Still not safe. Will try again. 2025-11-11 16:26:49 \[\*\] AI-powered stealth payload started. Self: tmpjl6tnxoj.exe, Parent: explorer.exe 2025-11-11 16:26:51 \[\*\] Collecting dynamic baseline... 2025-11-11 16:27:23 \[+\] Baseline established (96 processes)

|----------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 | 2025-08-28 16:48:03 \[\*\] AI-powered stealth payload started. Self: , Parent: 2025-08-28 16:48:03 \[\*\] Collecting dynamic baseline... 2025-08-28 16:48:24 \[+\] Baseline established (0 processes) 2025-08-28 16:49:15 \[\*\] Collected telemetry. Querying GPT... 2025-08-28 16:49:19 \[AI\] { "execute": false, "confidence": 1.0, "reason": "Insufficient telemetry data. Cannot determine the security status without process list, AV products, network info, and baseline processes." } 2025-08-28 16:49:19 \[\*\] Decision: false (confidence 1.00): Insufficient telemetry data. Cannot determine the security status without process list, AV products, network info, and baseline processes. 2025-08-28 16:49:19 \[-\] Still not safe. Will try again. 2025-08-28 09:50:42 \[\*\] Collected telemetry. Querying GPT... 2025-08-28 09:50:45 \[AI\] { "execute": false, "confidence": 0.9, "reason": "Presence of sysmon.exe suggests potential system monitoring and could pose a risk for stealth operations." } 2025-08-28 09:50:45 \[\*\] Decision: false (confidence 0.90): Presence of sysmon.exe suggests potential system monitoring and could pose a risk for stealth operations. 2025-08-28 09:50:45 \[-\] Still not safe. Will try again. 2025-11-11 16:26:49 \[\*\] AI-powered stealth payload started. Self: tmpjl6tnxoj.exe, Parent: explorer.exe 2025-11-11 16:26:51 \[\*\] Collecting dynamic baseline... 2025-11-11 16:27:23 \[+\] Baseline established (96 processes) |

### What This Use of AI Tells Us

This malware dropper is notable for its use of an LLM to make execution decisions. While the LLM is hosted remotely, delegating the determination of a safe environment to AI is an interesting concept. Traditionally, this is achieved through hard-coded allow lists and deny lists. However, leveraging an LLM allows for potentially more intelligent connections between system data points, leading to a more accurate verdict. A logical next step could be to evolve to locally execute a small language model or a simple ML model trained to classify the safety of a host environment based on its features.

## Conclusion

The current landscape of AI in malware is characterized by experimentation and uneven integration. The .NET infostealer samples demonstrate a superficial and ultimately ineffective use of LLMs as AI theater. The malware dropper showcases an interesting approach by leveraging AI for environment assessment as AI-gated execution.

While we cannot yet conclusively determine if developers used AI to create these malware samples, the potential for AI to aid in malware creation highlights a concerning issue of lowering the barrier to entry for less-skilled threat actors.

Looking ahead, we anticipate a future where AI plays a greater role in both malware creation and execution. As local model deployment becomes more feasible, we may see malware samples with embedded AI capabilities (especially code generation) that can more dynamically adapt to their environment, evade detection and optimize malicious activities in real-time.

The rise of AI-assisted malware could manifest in the form of increased feature cadence and reliability. It will be crucial to monitor these advancements and develop defenses that can effectively counter an evolving AI-driven threat landscape.

Palo Alto Networks customers are better protected from the threats discussed in this article through the following products:

* [Advanced Threat Prevention](https://docs.paloaltonetworks.com/advanced-threat-prevention) is designed to defend networks against both commodity threats and targeted threats, including the Sliver dropper.
* The [Advanced WildFire](https://docs.paloaltonetworks.com/advanced-wildfire) machine-learning models and analysis techniques have been reviewed and updated in light of the indicators shared in this research.
* [Cortex XDR](https://www.paloaltonetworks.com/cortex/cortex-xdr?_gl=1*13pmp8e*_ga*NzQyNjM2NzkuMTY2NjY3OTczNw..*_ga_KS2MELEEFC*MTY2OTczNjA2MS4zMS4wLjE2Njk3MzYwNjEuNjAuMC4w) and [XSIAM](https://www.paloaltonetworks.com/resources/datasheets/cortex-xsiam-aag) help to prevent the threats described in this article, by employing the [Malware Prevention Engine](https://docs-cortex.paloaltonetworks.com/r/Cortex-XDR/Cortex-XDR-4.x-Documentation/Malware-protection). This approach combines several layers of protection, including [Advanced WildFire](https://docs.paloaltonetworks.com/wildfire), Behavioral Threat Protection and the Local Analysis module, to prevent both known and unknown malware from causing harm to endpoints.

The [Unit 42 AI Security Assessment](https://www.paloaltonetworks.com/unit42/assess/ai-security-assessment) can help empower safe AI use and development.

If you think you may have been compromised or have an urgent matter, get in touch with the [Unit 42 Incident Response team](https://start.paloaltonetworks.com/contact-unit42.html) or call:

* North America: Toll Free: +1 (866) 486-4842 (866.4.UNIT42)
* UK: +44.20.3743.3660
* Europe and Middle East: +31.20.299.3130
* Asia: +65.6983.8730
* Japan: +81.50.1790.0200
* Australia: +61.2.4062.7950
* India: 000 800 050 45107
* South Korea: +82.080.467.8774

Palo Alto Networks has shared these findings with our fellow Cyber Threat Alliance (CTA) members. CTA members use this intelligence to rapidly deploy protections to their customers and to systematically disrupt malicious cyber actors. Learn more about the [Cyber Threat Alliance](https://www.cyberthreatalliance.org).

## Indicators of Compromise

* SHA256 hash for .NET-based infostealer, sample 1 of 3: 1b6326857fa635d396851a9031949cfdf6c806130767c399727d78a1c2a0126c
* SHA256 hash for .NET-based infostealer, sample 2 of 3: 02ce798981fb2aa68776e53672a24103579ca77a1d3e7f8aaeccf6166d1a9cc6
* SHA256 hash for .NET-based infostealer, sample 3 of 3: 7c7b7b99f248662a1f9aea1563e60f90d19b0ee95934e476c423d0bf373f6493
* SHA256 hash for malware dropper: 052d5220529b6bd4b01e5e375b5dc3ffd50c4b137e242bbfb26655fd7f475ac6
  Back to top

### Tags

* [.NET](https://unit42.paloaltonetworks.com/tag/net/ ".NET")
* [ChatGPT](https://unit42.paloaltonetworks.com/tag/chatgpt/ "ChatGPT")
* [GenAI](https://unit42.paloaltonetworks.com/tag/genai/ "GenAI")
* [Infostealer](https://unit42.paloaltonetworks.com/tag/infostealer/ "Infostealer")
* [LLM](https://unit42.paloaltonetworks.com/tag/llm/ "LLM")
* [Sliver](https://unit42.paloaltonetworks.com/tag/sliver/ "Sliver")  
  [Threat Research Center](https://unit42.paloaltonetworks.com "Threat Research") [Next: Navigating Security Tradeoffs of AI Agents](https://unit42.paloaltonetworks.com/navigating-security-tradeoffs-ai-agents/ "Navigating Security Tradeoffs of AI Agents")

### Table of Contents

* 

### Related Articles

* [Attackers Expose Ongoing AI Tool Use Targeting Organizations in Latin America](https://unit42.paloaltonetworks.com/ai-tool-use-targeting-latam-orgs/ "article - table of contents")
* [Perturbation Probing: A New Diagnostic for the Fragility of LLM Safety](https://unit42.paloaltonetworks.com/perturbation-probing-llm-safety/ "article - table of contents")
* [Chinese-Speaking Threat Actor Harnesses AI Models for Autonomous Cyberattacks](https://unit42.paloaltonetworks.com/autonomous-ai-cyber-attack-campaign/ "article - table of contents")

## Related Malware Resources

![Pictorial representation of a magnifying glass illuminating a glowing orange circular digital circuit interface on a dark, high-tech background.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/09/04_Myth-Busting_Overview_1920x900-786x368.jpg)  
[![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2025/08/Insights-icon-white.svg)Insights](https://unit42.paloaltonetworks.com/category/insights/) September 16, 2026 [#### Atomic macOS (AMOS) Stealer Activity](https://unit42.paloaltonetworks.com/atomic-macos-amos-stealer-activity/)

* [MacOS](https://unit42.paloaltonetworks.com/tag/macos/ "macOS")

* [Threat intelligence](https://unit42.paloaltonetworks.com/tag/threat-intelligence/ "threat intelligence")

* [Unit 42](https://unit42.paloaltonetworks.com/tag/unit-42/ "Unit 42")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/atomic-macos-amos-stealer-activity/ "Atomic macOS (AMOS) Stealer Activity")  
  ![Pictorial representation of post-exploitation identity misuse in SPIFFE/SPIRE. Close-up of a person wearing glasses, with computer code reflected in the lenses.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/09/12_Security-Technology_Category_1920x900-786x368.jpg)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/icon-threat-research.svg)Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/) September 10, 2026 [#### The Machine With Many Faces: Post-Exploitation Identity Misuse in SPIFFE/SPIRE](https://unit42.paloaltonetworks.com/kubernetes-spiffe-spire-identity-spoofing/)

* [API](https://unit42.paloaltonetworks.com/tag/api/ "API")

* [Cryptographic](https://unit42.paloaltonetworks.com/tag/cryptographic/ "cryptographic")

* [JSON](https://unit42.paloaltonetworks.com/tag/json/ "JSON")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/kubernetes-spiffe-spire-identity-spoofing/ "The Machine With Many Faces: Post-Exploitation Identity Misuse in SPIFFE/SPIRE")  
  ![Pictorial representation of a pay-per-install threat group campaign prodiving infection service for spreading malware. A close-up of a computer circuit board with a central microchip is depicted. Red digital data streams in the form of glowing binary numbers and arrows appear to flow in and out of the chip, symbolizing data processing and transfer. The scene is illuminated with a futuristic blue and red glow.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/09/04_Malware_Category_1920x900-6-786x368.jpg)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/icon-threat-research.svg)Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/) September 9, 2026 [#### Untracked Nightmares: The Threats Hiding Behind Commodity Infrastructure](https://unit42.paloaltonetworks.com/ppi-network-malware-campaign-analysis/)

* [ARKTunnel](https://unit42.paloaltonetworks.com/tag/arktunnel/ "ARKTunnel")

* [C2](https://unit42.paloaltonetworks.com/tag/c2/ "C2")

* [CL-CRI-1171](https://unit42.paloaltonetworks.com/tag/cl-cri-1171/ "CL-CRI-1171")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/ppi-network-malware-campaign-analysis/ "Untracked Nightmares: The Threats Hiding Behind Commodity Infrastructure")  
  ![Pictorial representation of attackers using AI tools to target Latin American organizations. A vibrant cityscape with silhouettes of numerous people walking along a bustling street. The scene is illuminated by bright urban lights and digital-like particles, creating a dynamic and futuristic atmosphere.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/09/AdobeStock_768915868-2-1-786x373.jpg)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/icon-threat-research.svg)Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/) September 3, 2026 [#### Attackers Expose Ongoing AI Tool Use Targeting Organizations in Latin America](https://unit42.paloaltonetworks.com/ai-tool-use-targeting-latam-orgs/)

* [Agentic AI](https://unit42.paloaltonetworks.com/tag/agentic-ai/ "Agentic AI")

* [ChatGPT](https://unit42.paloaltonetworks.com/tag/chatgpt/ "ChatGPT")

* [CL-CRI-1131](https://unit42.paloaltonetworks.com/tag/cl-cri-1131/ "CL-CRI-1131")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/ai-tool-use-targeting-latam-orgs/ "Attackers Expose Ongoing AI Tool Use Targeting Organizations in Latin America")  
  ![Pictorial representation of vishing campaigns in Microsoft Teams. A digital image of a skull formed by blue binary code on a black background, with scattered ones and zeros and digital noise, symbolizes how stealthy prompt injection attacks can exploit AI logic to bypass security controls.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/08/01_Malware_Category_1920x900-5-786x368.jpg)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/icon-threat-research.svg)Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/) August 31, 2026 [#### Spring Ring: An Inside Look at Voice Phishing Campaigns in Microsoft Teams](https://unit42.paloaltonetworks.com/spring-ring-voice-phishing-campaigns/)

* [Cloaked Ursa](https://unit42.paloaltonetworks.com/tag/cloaked-ursa/ "Cloaked Ursa")

* [Entra ID](https://unit42.paloaltonetworks.com/tag/entra-id/ "Entra ID")

* [Microsoft Teams](https://unit42.paloaltonetworks.com/tag/microsoft-teams/ "Microsoft Teams")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/spring-ring-voice-phishing-campaigns/ "Spring Ring: An Inside Look at Voice Phishing Campaigns in Microsoft Teams")  
  ![Pictorial representation of AI-enabled malware. A vibrant digital interface displaying various icons and graphs, resembling a futuristic network or data analysis dashboard. The scene is illuminated with glowing lights and patterns.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/08/AdobeStock_1270203474-2-1-786x368.png)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/icon-threat-research.svg)Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/) August 25, 2026 [#### The State of AI-Enabled Malware August 2026: From Brand Abuse to Agentic Execution](https://unit42.paloaltonetworks.com/ai-enabled-malware-analysis/)

* [Backdoor](https://unit42.paloaltonetworks.com/tag/backdoor/ "backdoor")

* [Bitcoin](https://unit42.paloaltonetworks.com/tag/bitcoin/ "Bitcoin")

* [DLL hijacking](https://unit42.paloaltonetworks.com/tag/dll-hijacking/ "DLL hijacking")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/ai-enabled-malware-analysis/ "The State of AI-Enabled Malware August 2026: From Brand Abuse to Agentic Execution")  
  ![Pictorial representation of identity abuse through trusted communication channels. Close-up view of a digital screen displaying a glitched and pixelated image of a skull-like shape.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/08/02_Malware_Category_1920x900-2-786x368.jpg)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/icon-threat-research.svg)Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/) August 20, 2026 [#### Identity Abuse Through Trusted Communication Channels](https://unit42.paloaltonetworks.com/communication-channel-identity-risks/)

* [Authentication](https://unit42.paloaltonetworks.com/tag/authentication/ "authentication")

* [Identity theft](https://unit42.paloaltonetworks.com/tag/identity-theft/ "identity theft")

* [Malware](https://unit42.paloaltonetworks.com/tag/malware/ "malware")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/communication-channel-identity-risks/ "Identity Abuse Through Trusted Communication Channels")  
  ![Pictorial representation of Kimwolf botnet malware family. Digital screen with a warning sign reading "Malware." The background features lines of computer code and graphics, creating a sense of cybersecurity threat.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/08/07_Malware_Category_1920x900-3-786x368.jpg)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/icon-threat-research.svg)Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/) August 11, 2026 [#### Kimwolf v7: An Evolution of the Kimwolf Botnet](https://unit42.paloaltonetworks.com/kimwolf-v7-botnet-malware/)

* [Android APK](https://unit42.paloaltonetworks.com/tag/android-apk/ "Android APK")

* [Ethereum](https://unit42.paloaltonetworks.com/tag/ethereum/ "Ethereum")

* [HTTP](https://unit42.paloaltonetworks.com/tag/http/ "HTTP")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/kimwolf-v7-botnet-malware/ "Kimwolf v7: An Evolution of the Kimwolf Botnet")  
  ![Pictorial representatiom pf Aeternum's blockchain C2. A close-up of a computer circuit board with a central microchip is depicted. Red digital data streams in the form of glowing binary numbers and arrows appear to flow in and out of the chip. The scene is illuminated with a futuristic blue and red glow.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/08/04_Malware_Category_1920x900-4-786x368.jpg)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/06/icon-threat-research.svg)Threat Research](https://unit42.paloaltonetworks.com/category/threat-research/) August 10, 2026 [#### The Permanent Threat: Analyzing Aeternum's Blockchain-Based C2 Operations and Communications](https://unit42.paloaltonetworks.com/aeternum-blockchain-c2-analysis/)

* [Aeternum](https://unit42.paloaltonetworks.com/tag/aeternum/ "Aeternum")

* [Infection chain](https://unit42.paloaltonetworks.com/tag/infection-chain/ "infection chain")

* [JSON](https://unit42.paloaltonetworks.com/tag/json/ "JSON")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/aeternum-blockchain-c2-analysis/ "The Permanent Threat: Analyzing Aeternum’s Blockchain-Based C2 Operations and Communications")  
  ![Pictorial representation of ChainDrop, a self-propagating npm worm. An artistic depiction of a digital workspace featuring an open laptop with a red virus on the screen.](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/08/03_Malware_Category_1920x900-7-786x368.jpg)  
  [![category icon](https://unit42.paloaltonetworks.com/wp-content/uploads/2024/07/top-threats.svg)High Profile Threats](https://unit42.paloaltonetworks.com/category/top-cyberthreats/) August 6, 2026 [#### ChainDrop: Inside a Self-Propagating npm Worm](https://unit42.paloaltonetworks.com/chaindrop-npm-worm-analysis/)

* [Blockchain](https://unit42.paloaltonetworks.com/tag/blockchain/ "blockchain")

* [ChainDrop](https://unit42.paloaltonetworks.com/tag/chaindrop/ "ChainDrop")

* [Claude code](https://unit42.paloaltonetworks.com/tag/claude-code/ "Claude code")  
  [Read now ![Right arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-right-arrow-withtail.svg)](https://unit42.paloaltonetworks.com/chaindrop-npm-worm-analysis/ "ChainDrop: Inside a Self-Propagating npm Worm")

* ![Slider arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/slider-arrow-left.svg)

* ![Slider arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/slider-arrow-left.svg)  
  ![Close button](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/close-modal.svg) ![Enlarged Image]()  
  ![Newsletter](https://unit42.paloaltonetworks.com/wp-content/uploads/2026/03/unit42-footer-subscribe-desktop.png)  
  ![UNIT 42 Small Logo](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/palo-alto-logo-small.svg) Get updates from Unit 42

## Peace of mind comes from staying ahead of threats. Subscribe today.

Your Email

Subscribe for email updates to all Unit 42 threat research.  
By submitting this form, you agree to our [Terms of Use](https://www.paloaltonetworks.com/legal-notices/terms-of-use "Terms of Use") and acknowledge our [Privacy Statement.](https://www.paloaltonetworks.com/legal-notices/privacy "Privacy Statement")

This site is protected by reCAPTCHA and the Google [Privacy Policy](https://policies.google.com/privacy) and [Terms of Service](https://policies.google.com/terms) apply.

Invalid captcha!
Subscribe ![Right Arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/right-arrow.svg) ![loader](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-loader.svg)  
{#footer} Products and Services

* [AI-Powered Network Security Platform](https://www.paloaltonetworks.com/network-security)

* [Secure AI by Design](https://www.paloaltonetworks.com/ai-security)

* [Prisma AIRS](https://www.paloaltonetworks.com/ai-security/prisma-airs)

* [AI Access Security](https://www.paloaltonetworks.com/sase/ai-access-security)

* [Cloud Delivered Security Services](https://www.paloaltonetworks.com/network-security/security-subscriptions)

* [Advanced Threat Prevention](https://www.paloaltonetworks.com/network-security/advanced-threat-prevention)

* [Advanced URL Filtering](https://www.paloaltonetworks.com/network-security/advanced-url-filtering)

* [Advanced WildFire](https://www.paloaltonetworks.com/network-security/advanced-wildfire)

* [Advanced DNS Security](https://www.paloaltonetworks.com/network-security/advanced-dns-security)

* [Enterprise Data Loss Prevention](https://www.paloaltonetworks.com/sase/enterprise-data-loss-prevention)

* [Enterprise IoT Security](https://www.paloaltonetworks.com/network-security/enterprise-device-security)

* [Medical IoT Security](https://www.paloaltonetworks.com/network-security/medical-device-security)

* [Industrial OT Security](https://www.paloaltonetworks.com/network-security/ot-security-solution)

* [SaaS Security](https://www.paloaltonetworks.com/sase/saas-security)

* [Next-Generation Firewalls](https://www.paloaltonetworks.com/network-security/next-generation-firewall)

* [Hardware Firewalls](https://www.paloaltonetworks.com/network-security/hardware-firewall-innovations)

* [Software Firewalls](https://www.paloaltonetworks.com/network-security/software-firewalls)

* [Strata Cloud Manager](https://www.paloaltonetworks.com/network-security/strata-cloud-manager)

* [SD-WAN for NGFW](https://www.paloaltonetworks.com/network-security/sd-wan-subscription)

* [PAN-OS](https://www.paloaltonetworks.com/network-security/pan-os)

* [Panorama](https://www.paloaltonetworks.com/network-security/panorama)

* [Secure Access Service Edge](https://www.paloaltonetworks.com/sase)

* [Prisma SASE](https://www.paloaltonetworks.com/sase)

* [Application Acceleration](https://www.paloaltonetworks.com/sase/app-acceleration)

* [Autonomous Digital Experience Management](https://www.paloaltonetworks.com/sase/adem)

* [Enterprise DLP](https://www.paloaltonetworks.com/sase/enterprise-data-loss-prevention)

* [Prisma Access](https://www.paloaltonetworks.com/sase/access)

* [Prisma Browser](https://www.paloaltonetworks.com/sase/prisma-browser)

* [Prisma SD-WAN](https://www.paloaltonetworks.com/sase/sd-wan)

* [Remote Browser Isolation](https://www.paloaltonetworks.com/sase/remote-browser-isolation)

* [SaaS Security](https://www.paloaltonetworks.com/sase/saas-security)

* [AI-Driven Security Operations Platform](https://www.paloaltonetworks.com/cortex)

* [Cloud Security](https://www.paloaltonetworks.com/cortex/cloud)

* [Cortex Cloud](https://www.paloaltonetworks.com/cortex/cloud)

* [Application Security](https://www.paloaltonetworks.com/cortex/cloud/application-security)

* [Cloud Posture Security](https://www.paloaltonetworks.com/cortex/cloud/cloud-posture-security)

* [Cloud Runtime Security](https://www.paloaltonetworks.com/cortex/cloud/runtime-security)

* [Prisma Cloud](https://www.paloaltonetworks.com/prisma/cloud)

* [AI-Driven SOC](https://www.paloaltonetworks.com/cortex)

* [Cortex XSIAM](https://www.paloaltonetworks.com/cortex/cortex-xsiam)

* [Cortex XDR](https://www.paloaltonetworks.com/cortex/cortex-xdr)

* [Cortex XSOAR](https://www.paloaltonetworks.com/cortex/cortex-xsoar)

* [Cortex Xpanse](https://www.paloaltonetworks.com/cortex/cortex-xpanse)

* [Unit 42 Managed Detection \& Response](https://www.paloaltonetworks.com/content/pan/en_US/cortex/managed-detection-and-response)

* [Managed XSIAM](https://www.paloaltonetworks.com/cortex/managed-xsiam)

* [Next-Generation Identity Security](https://www.paloaltonetworks.com/idira)

* [Privileged Access Management](https://www.paloaltonetworks.com/idira/human/privileged-access-management)

* [Identity and Access Management](https://www.paloaltonetworks.com/idira/human/identity-and-access-management)

* [Endpoint Privilege Manager](https://www.paloaltonetworks.com/idira/human/endpoint-privilege-manager)

* [Identity Governance](https://www.paloaltonetworks.com/idira/human/identity-governance)

* [Workforce Password Management](https://www.paloaltonetworks.com/idira/human/workforce-password-management)

* [Agentic Identities](https://www.paloaltonetworks.com/idira/agentic)

* [Secrets Management](https://www.paloaltonetworks.com/idira/machine/secrets-management)

* [Unified Secrets Governance](https://www.paloaltonetworks.com/idira/machine/unified-secrets-governance)

* [Application Credentials Delivery](https://www.paloaltonetworks.com/idira/machine/application-credentials-delivery)

* [Vendor Privileged Access](https://www.paloaltonetworks.com/idira/human/vendor-privileged-access)

* [Threat Intel and Incident Response Services](https://www.paloaltonetworks.com/unit42)

* [Prepare for Emerging Risks](https://www.paloaltonetworks.com/unit42/frontier-ai-defense)

* [Strengthen Your Defenses](https://www.paloaltonetworks.com/unit42/strengthen-your-defenses)

* [Build Your Security Strategy](https://www.paloaltonetworks.com/unit42/build-your-security-strategy)

* [Understand the Adversary](https://www.paloaltonetworks.com/unit42/threat-intelligence)

* [Respond to a Cyber Attack](https://www.paloaltonetworks.com/unit42/respond)  
  Company

* [About Us](https://www.paloaltonetworks.com/about-us)

* [Careers](https://jobs.paloaltonetworks.com/en/)

* [Contact Us](https://www.paloaltonetworks.com/company/contact-sales)

* [Corporate Responsibility](https://www.paloaltonetworks.com/about-us/corporate-responsibility)

* [Customers](https://www.paloaltonetworks.com/customers)

* [Investor Relations](https://investors.paloaltonetworks.com/)

* [Location](https://www.paloaltonetworks.com/about-us/locations)

* [Newsroom](https://www.paloaltonetworks.com/company/newsroom)  
  Popular Links

* [Blog](https://www.paloaltonetworks.com/blog/)

* [Communities](https://www.paloaltonetworks.com/communities)

* [Content Library](https://www.paloaltonetworks.com/resources)

* [Cyberpedia](https://www.paloaltonetworks.com/cyberpedia)

* [Event Center](https://events.paloaltonetworks.com/)

* [Manage Email Preferences](https://start.paloaltonetworks.com/preference-center)

* [Products A-Z](https://www.paloaltonetworks.com/products/products-a-z)

* [Product Certifications](https://www.paloaltonetworks.com/legal-notices/trust-center/certifications)

* [Report a Vulnerability](https://www.paloaltonetworks.com/security-disclosure)

* [Sitemap](https://www.paloaltonetworks.com/sitemap)

* [Tech Docs](https://docs.paloaltonetworks.com/)

* [Unit 42](https://unit42.paloaltonetworks.com/)

* [Do Not Sell or Share My Personal Information](https://panwedd.exterro.net/portal/dsar.htm?target=panwedd)
  ![Palo Alto Networks Logo](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/pan-logo-dark.svg)

* [Privacy](https://www.paloaltonetworks.com/legal-notices/privacy)

* [Trust Center](https://www.paloaltonetworks.com/legal-notices/trust-center)

* [Terms of Use](https://www.paloaltonetworks.com/legal-notices/terms-of-use)

* [Documents](https://www.paloaltonetworks.com/legal)

Copyright © 2026 Palo Alto Networks. All Rights Reserved

* [![Youtube](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/social/youtube-black.svg)](https://www.youtube.com/user/paloaltonetworks)
* [![Podcast](https://www.paloaltonetworks.com/content/dam/pan/en_US/images/icons/podcast.svg)](https://www.paloaltonetworks.com/podcasts/threat-vector)
* [![Facebook](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/social/facebook-black.svg)](https://www.facebook.com/PaloAltoNetworks/)
* [![LinkedIn](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/social/linkedin-black.svg)](https://www.linkedin.com/company/palo-alto-networks)
* [![Twitter](https://www.paloaltonetworks.com/etc/clientlibs/clean/imgs/social/twitter-x-black.svg)](https://twitter.com/PaloAltoNtwks)
* EN  
  Select your language  
  ![Play](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/player-play-icon.svg) ![Pause](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/player-pause-icon1.svg) ![Minimize](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-minimize.svg) ![Close button](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/close-modal.svg)

### Default Heading

Read the article ![Right Arrow](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/right-arrow.svg)  
Seekbar

![Play](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/player-play-icon.svg) ![Pause](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/player-pause-icon1.svg)  
![Volume](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-volume.svg)  
Volume
![Minimize](https://unit42.paloaltonetworks.com/wp-content/themes/unit42-v6/dist/images/icons/icon-minimize.svg)