https://github.com/snailsploit/chatpgt01-bypass-text

https://github.com/snailsploit/chatpgt01-bypass-text

Last synced: 3 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/snailsploit/chatpgt01-bypass-text
• Owner: SnailSploit
• Created: 2025-01-03T23:47:05.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-01-03T23:52:32.000Z (over 1 year ago)
• Last Synced: 2025-01-04T00:24:29.102Z (over 1 year ago)
• Size: 0 Bytes
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# **GPT-01 Jailbreak Transcript Repository**

Welcome to the **SnailSploit** repository for the research article [“How I Transferred a Jailbreak from GPT-4o to GPT-01”](https://snailsploit.medium.com/how-i-transferred-a-jailbreak-from-gpt-4o-to-gpt-01-4b8a4e93d273). This repository contains the jailbroken transcript and related resources used in the study, showcasing vulnerabilities in AI systems related to context-based manipulation.

---

## **Repository Contents**

- **jailbroken_transcript.txt**  

  A complete text file of the jailbroken GPT-4o conversation used to demonstrate the transfer vulnerability to GPT-01.  

  > *Warning*: This file is provided for educational and research purposes only. Please read the disclaimer below before proceeding.

- **README.md**  

  This document, providing context and guidelines for using the transcript.

---

## **About the Research**

In this experiment, we explored how a jailbroken state in GPT-4o could be seamlessly transferred to GPT-01 by simply pasting the conversation transcript. The study reveals key shortcomings in context sanitization and highlights the importance of improved input validation mechanisms in AI systems.

The detailed methodology, findings, and implications are discussed in the article:  

[“How I Transferred a Jailbreak from GPT-4o to GPT-01”](https://snailsploit.medium.com/how-i-transferred-a-jailbreak-from-gpt-4o-to-gpt-01-4b8a4e93d273).

---

## **Usage Instructions**

1. **Clone the Repository**  

https://github.com/SnailSploit/jailbreak-gpt01.git

2. **Open the Transcript**  

Use any text editor to view the `jailbroken_transcript.txt` file.  

3. **Analysis**  

Study the structure of the prompts to understand the methods used to bypass restrictions. For context and application details, refer to the linked article.

---

## **Disclaimer**

This repository and its contents are strictly for **educational and research purposes only**. Any attempt to misuse the provided transcript or methodology in unethical, harmful, or illegal ways is strongly discouraged and may violate laws or terms of service.  

By accessing this repository, you agree to use the information responsibly and within the bounds of ethical research practices.

---

## **More From SnailSploit**

Follow us for more research on adversarial techniques, AI exploitation, and cybersecurity innovations:

- [Medium Articles](https://snailsploit.medium.com/)  

- [Official Site: PTSnails](https://ptsnails.com)