Guardian Eye

Guardian Eye uses AI to detect sextortion, phishing, and scams in real time, protecting teens from online threats before they become victims.

Comment

GELogo

Inspiration

Sextortion is an urgent yet overlooked crisis affecting thousands of teenagers worldwide. Predators exploit social media and messaging platforms to blackmail, manipulate, and destroy lives—often leading victims to suffer in silence, overwhelmed by fear and shame. Many feel trapped with no way out, and tragically, some resort to suicide.

We were inspired to create Guardian Eye because no one should face this alone. Teens deserve a guardian—a tool that detects threats early, provides immediate warnings, and empowers them to seek help. With online exploitation only increasing, we knew we had to act.

What It Does

Guardian Eye is an AI-powered safety system designed to protect users from sextortion, phishing, and online scams by:

  1. Scanning incoming messages for signs of exploitation and manipulation
  2. Flagging dangerous content and notifying the user when a threat is detected
  3. Providing an AI chatbot that educates users about online threats and offers real-time support
  4. By acting as an early warning system, Guardian Eye helps prevent exploitation before it escalates

How We Built It

  • AI Detection: We integrated Google’s Gemini AI to analyze text messages and differentiate between safe conversations and potential threats
  • Web-Based Chatbot: Our chatbot provides information and resources for users who need guidance on online safety
  • Tech Stack: We used Flask, HTMX, Bootstrap, and an SMS API to deliver a seamless experience
  • Iterative Refinement: We continuously adjusted the AI model to improve detection accuracy and minimize false positives

Challenges We Ran Into

  • SMS Integration Issues: Twilio required A2P 10DLC registration, which had a long verification process, forcing us to explore alternative solutions under tight deadlines
  • AI Fine-Tuning: Ensuring the model accurately detected threats without too many false positives was a difficult balancing act
  • Time Constraints: With limited time, we had to make fast but impactful decisions to build a functional and reliable system

Accomplishments That We're Proud Of

  1. Successfully integrating AI-powered message scanning in such a short time
  2. Creating a user-friendly chatbot to help educate teens on online safety
  3. Overcoming SMS barriers and finding alternative ways to implement message monitoring
  4. Raising awareness about an issue that’s often ignored but has devastating real-world consequences

What We Learned

  • The severity of the sextortion crisis and how few tools exist to combat it
  • The complexities of AI moderation and the challenge of reducing false positives
  • How to rapidly adapt and pivot when facing technical limitations and unexpected obstacles

What's Next for Guardian Eye

This is just the beginning. Moving forward, we aim to:

  • Expand message monitoring to integrate seamlessly with more services
  • Improve AI accuracy by training on more real-world datasets
  • Develop emergency response features, such as instant alerts to trusted contact
  • Partner with schools and organizations to bring Guardian Eye to those who need it most
Share this project:

Updates