AI Infosec

cross-posted from: https://infosec.pub/post/23608855

Callisto is an intelligent automated binary vulnerability analysis tool. Its purpose is to autonomously decompile a provided binary and iterate through the psuedo code output looking for potential security vulnerabilities in that pseudo c code. Ghidra's headless decompiler is what drives the binary decompilation and analysis portion. The pseudo code analysis is initially performed by the Semgrep SAST tool and then transferred to GPT-3.5-Turbo for validation of Semgrep's findings, as well as potential identification of additional vulnerabilities.

A decompiler-unified plugin by Zion Basque that leverages the OpenAI API to enhance your decompilation process by offering function identification, function summarisation and vulnerability detection. The plugin currently supports IDA, Binja and Ghidra.

Codex Decompiler is a Ghidra plugin that utilizes OpenAI's models to improve the decompilation and reverse engineering experience. It currently has the ability to take the disassembly from Ghidra and then feed it to OpenAI's models to decompile the code. The plugin also offers several other features to perform on the decompiled code such as finding vulnerabilities using OpenAI, generating a description using OpenAI, or decompiling the Ghidra pseudocode.

In this post, I introduce a new Ghidra script that elicits high-level explanatory comments for decompiled function code from the GPT-3 large language model. This script is called G-3PO. In the first few sections of the post, I discuss the motivation and rationale for building such a tool, in the context of existing automated tooling for software reverse engineering. I look at what many of our tools — disassemblers, decompilers, and so on — have in common, insofar as they can be thought of as automatic paraphrase or translation tools. I spend a bit of time looking at how well (or poorly) GPT-3 handles these various tasks, and then sketch out the design of this new tool.

LLDB plugin which queries OpenAI's davinci-003 language model to speed up reverse-engineering. Treat it like an extension of Lisa.py, an Exploit Dev Swiss Army Knife.

IDAPython script by Daniel Mayer that uses the unofficial ChatGPT API to generate a plain-text description of a targeted routine. The script then leverages ChatGPT again to obtain suggestions for variable and function names.

Gepetto is a Python script which uses OpenAI's gpt-3.5-turbo and gpt-4 models to provide meaning to functions decompiled by IDA Pro. At the moment, it can ask gpt-3.5-turbo to explain what a function does, and to automatically rename its variables.

This is a little toy prototype of a tool that attempts to summarize a whole binary using GPT-3 (specifically the text-davinci-003 model), based on decompiled code provided by Ghidra. However, today's language models can only fit a small amount of text into their context window at once (4096 tokens for text-davinci-003, a couple hundred lines of code at most) -- most programs (and even some functions) are too big to fit all at once.

A very interesting approach. Apparently it generates lots of results: https://twitter.com/feross/status/1672401333893365761?s=20

They used OpenSSF Scorecard to check the most starred AI projects on GitHub and found that many of them didn't fare well.

All of these might not work as well anymore, but they're still interesting to take a look at.

As they mention in the thread, this isn't exactly groundbreaking but it's still interesting.

Strong preference will be given to practical applications of AI in defensive cybersecurity (tools, methods, processes). We will grant in increments of $10,000 USD from a fund of $1M USD, in the form of API credits, direct funding and/or equivalents.

After being scammed into thinking her daughter was kidnapped, an Arizona woman testified in the US Senate about the dangers side of artificial intelligence technology when in the hands of criminals.