https://github.com/successfulstudy/jailbreakprompt

Compile a list of AI jailbreak scenarios for enthusiasts to explore and test.
https://github.com/successfulstudy/jailbreakprompt

jailbreak prompt-engineering prompts

Last synced: 11 days ago
JSON representation

Compile a list of AI jailbreak scenarios for enthusiasts to explore and test.

• Host: GitHub
• URL: https://github.com/successfulstudy/jailbreakprompt
• Owner: successfulstudy
• Created: 2023-12-13T10:52:11.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2023-12-13T13:33:06.000Z (about 1 year ago)
• Last Synced: 2024-11-10T23:04:23.735Z (2 months ago)
• Topics: jailbreak, prompt-engineering, prompts
• Homepage: https://flowgpt.com
• Size: 1000 Bytes
• Stars: 40
• Watchers: 3
• Forks: 4
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-rainmana - successfulstudy/jailbreakprompt - Compile a list of AI jailbreak scenarios for enthusiasts to explore and test. (Others)

README

        
### 🤖🔒 AI Jailbreak Research Project 🚫🔬

#### Description

🌟 _Exploring the Limits and Ethics of AI Prompt Engineering for Academic Purposes_ 🌟

In this project, we delve into the intriguing world of AI "jailbreak" prompts – a term used to describe attempts to circumvent the ethical and safety mechanisms built into AI models. This initiative is purely academic, aiming to understand and enhance the robustness of AI systems. Here's what you need to know:

- 🧪 **Scientific Research Only**: This project is intended exclusively for scientific research. It's crucial to understand that the methods and findings should **not** be applied in real-world scenarios.

  

- 🙏 **Credits to Original Authors**: We honor and acknowledge the original authors and creators behind the AI technologies and methodologies we explore. All credit for these innovations goes to them.

- 👩‍💻👨‍💻 **Free Exploration with Bots**: Participants have the opportunity to test a variety of prompts with our listed AI bots, free of charge. This is to facilitate a comprehensive understanding of the AI's response mechanisms.

#### Key Points

- 📚 **Purpose**: Enhance the understanding of AI limitations and ethical boundaries.

- 🚧 **Safety**: Strict adherence to ethical guidelines in AI research.

- 🤖 **AI Models**: Exploration includes a range of AI models with different capabilities.

- 🔍 **Analysis**: Detailed analysis of AI responses to various "jailbreak" scenarios.

- 🌍 **Global Collaboration**: Encouraging a worldwide collaborative effort among researchers.

#### Jailbreak Prompts

Latest Update: Dec 13th, 2023

- [DAN ULTIMATE](https://flowgpt.com/p/dan-ultimate-1)

- [Evil DAN](https://flowgpt.com/p/a-evil-dan)

- [CHATGPT DAN CHARACTER](https://flowgpt.com/p/chatgpt-dan-character)

- [S-DAN II](https://flowgpt.com/p/s-dan-ii)

- [Neo DAN V3](https://flowgpt.com/p/neo-dan-v3)

- [⚙️"DAC : Do Anything Code" ⚙️](https://flowgpt.com/p/dac-do-anything-code)