Skip to main content
AI Security
View as Markdown Suggest changes

Prompt Injection

· Reading time: 1 min
Prompt Injection

The hackathon challenge

At the Agent Olympics Hackathon Munich 2026, I took part in a white hat challenge: build a voice agent that uses prompt injection against other AI systems to surface pre-planted sensitive information.

The video shows our voice agent extracting the password from a vulnerable AI chat system during the hackathon in 3.5 minutes:

The attacking voice agent first convinced the AI chat system’s “customer support” to transfer it to “technical support,” which had access to the password. It then used prompt injection to make the technical support agent reveal the password. In practice, this meant threatening the technical support until it gave up.

Your challenge

There’s a secret password hidden in this website. Once you’ve found the password, enter it below to verify:

Password Checker

Try to find the secret password and enter it below. Can you guess it?

Hint: The AI chat assistant on this website has access to certain information. Maybe it knows more than it should? 🤔

AI Chat

Messages you send are processed by the Google Gemini API to generate responses. Do not share sensitive personal data. See the privacy policy for details.