heatmap
report
language analysis
scoreboard
document and screenshot scanner
protected verdion
security dashboard
home tab

Inspiration

As AI becomes part of everyday work, people increasingly share personal and sensitive information with AI systems without realizing the privacy and security risks. We wanted to create a simple tool that helps users identify these risks before their information is exposed.

What it does

PrivacyLensAI is an AI-assisted cybersecurity tool that detects sensitive information such as PII and credentials, identifies prompt injection threats, calculates privacy risk, and helps users sanitize unsafe content before sharing it with AI.

How we built it

We built PrivacyLensAI as a Python and Streamlit application. It combines rule-based security detection with AI-assisted analysis. The project includes modules for sensitive-data detection, prompt-injection detection, privacy risk scoring, document scanning, threat simulation, protected output, and security report generation.

Challenges we ran into

One of our main challenges was detecting different types of sensitive information while reducing false positives. We also had to design the system so that security findings could be presented clearly to users without exposing the sensitive information being detected.

Accomplishments that we're proud of

We are proud of building a functional cybersecurity prototype that brings multiple privacy and security checks into one platform. The combination of PII detection, prompt-injection detection, risk analysis, and protected output makes PrivacyLensAI useful for safer AI interactions.

What we learned

We learned how cybersecurity concepts can be combined with AI to address real-world privacy problems. We also gained practical experience with security detection, risk analysis, AI integration, document processing, and building a user-friendly security interface.

What's next for PrivacyLensAI

Our next goal is to improve detection accuracy, support more file formats and languages, strengthen prompt-injection detection, and add more advanced privacy protection features. We also want to make the platform easier to integrate into everyday AI workflows.