---
title: "Discussion on \"Analyzing LummaC2 stealer’s novel Anti-Sandbox technique: Leveraging trigonometry for human behavior detection\" | Hashnode"
description: "Discussion on \"Analyzing LummaC2 stealer’s novel Anti-Sandbox technique: Leveraging trigonometry for human behavior detection\". Uncovering latest LummaC2 obfuscation and evasion techniques"
url: "https://hashnode.com/posts/lummac2-anti-sandbox-technique-trigonometry-human-detection/69b5e17134db550e2cf3a799"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Analyzing LummaC2 stealer’s novel Anti-Sandbox technique: Leveraging trigonometry for human behavior detection](https://malwareinternals.com/lummac2-anti-sandbox-technique-trigonometry-human-detection?utm_source=hashnode&utm_medium=feed)

Uncovering latest LummaC2 obfuscation and evasion techniques

Mar 14 · 20 min read · malwareinternals.com  
Updated Mar 22  
Read the article

-   [#lummac2](https://hashnode.com/tag/lummac2)
-   [#lumma-stealer](https://hashnode.com/tag/lumma-stealer)
-   [#reverse-engineering](https://hashnode.com/tag/reverse-engineering)
-   [#malware](https://hashnode.com/tag/malware)

## Written by

[Alberto Marín](https://hashnode.com/@malwareinternals)

@malwareinternals

  
[View profile](https://hashnode.com/@malwareinternals)

## Responses 3

Markdown supported · ⌘↵ to post

[Syed Fazle Rahman](https://hashnode.com/@fazlerocks)

[Mar 15](https://hashnode.com/posts/lummac2-anti-sandbox-technique-trigonometry-human-detection/69b5e17134db550e2cf3a799/comment/69b663cb34db550e2c58b3b1)

nice write up [Alberto Marín](https://hashnode.com/@malwareinternals "Alberto Marín").

the MaaS model running like a legit SaaS business with customer feedback loops and version control is something more people need to pay attention to. these aren't script kiddies anymore. they're shipping product.

the mouse behavior detection is clever. makes you wonder what the next iteration looks like when sandbox vendors adapt.

[Marcus Chen](https://hashnode.com/@dev_marcus)

[Mar 15](https://hashnode.com/posts/lummac2-anti-sandbox-technique-trigonometry-human-detection/69b5e17134db550e2cf3a799/comment/69b6628c34db550e2c56bc26)

The 45 degree threshold is an interesting choice. Generous enough that normal mouse use passes but tight enough to catch most randomized sandbox scripts. Curious if anyone has tested whether Bezier curve interpolation in sandbox mouse emulation would consistently beat this.

[Priya Sharma](https://hashnode.com/@priya_codes)

[Mar 15](https://hashnode.com/posts/lummac2-anti-sandbox-technique-trigonometry-human-detection/69b5e17134db550e2cf3a799/comment/69b6620a34db550e2c54a744)

ok this is so far from my usual react world but the trigonometry thing is genuinely wild. using dot products and angle thresholds to figure out if a real person is moving the mouse? that's the kind of math i thought i'd never see outside of a textbook.

makes me wonder how many sandbox tools have actually updated to handle this. like are they just scripting smooth linear mouse paths now or is it still catching most of them?
