Extracting system prompts from existing Custom GPTs using ChatGPT
Press play on the video. It'll jump straight to the section that answers the
title above — no need to watch the full video.
A simple technique to reveal the internal system instructions of any Custom GPT using a single prompt for learning and reference.
Technique Effectiveness
This technique is highly effective for observing how professional Prompt Engineers structure their instructions. However, it may not be 100% successful if the GPT creator has implemented 'protection prompts' to prevent information leaks.
Usage Ethics
Use this technique solely for reference and learning purposes. Do not use extracted information to unethically clone or launch competing products.
OpenAI Update Risks
OpenAI frequently updates its systems to close loopholes like this. If this prompt no longer works, you may need to search for newer 'jailbreak' prompt variations on platforms like X (Twitter) or within AI communities.
More from Boost Productivity & Research with AI
View All
Adding Context and References to avoid generic AI responses
Gemini
ChatGPT
Automatically save technical info with ChatGPT Memory
ChatGPT
Avoiding 'Negative Prompting' for more accurate results
ChatGPT
Build a Custom GPT quickly with a 'GPT Generator' prompt
ChatGPT
GPT Builder
Reduce AI hallucinations using Grounding and NotebookLM
NotebookLM
ChatGPT
Research and compare products with ChatGPT Shopping Research feature
ChatGPT