Use AI-generated malware and compress it using different settings. Analyze if AI-generated, packed code has a distinct Shannon entropy or LZma structure compared to human-written code.
Which direction sounds most interesting to you? I can help you outline the , abstract , or key findings for any of these. i dove down the 7z rabbit hole (it goes deep)
A new architecture for decentralized software distribution that reduces reliance on DNS trust, making it nearly impossible to trick users into downloading malicious installers. neil.7z
) can be used to hide malicious payloads that trigger buffer overflows in decompression routines.
Based on the technical, modern, and slightly chaotic context of "neil.7z" (referencing 7-Zip, compression, and potential security issues), here are three interesting, high-impact paper concepts: 1. Security/Cryptography Focus Use AI-generated malware and compress it using different
"The Ghost in the Archive: Detecting AI-Generated Malware via 7z Entropy Analysis"
"Trusting the Mirror: An Analysis of Malvertising in Software Repositories" I can help you outline the , abstract
A framework for detecting malicious code by analyzing the "difficulty" of compressing it, using compression ratios as a feature for security scanners. 3. Practical/Software Engineering Focus