Loading PhishGuard

Initializing AI model and security protocols...

Loading model: 0%

PHISHGUARD

Human vs AI Phishing Detection Research Study

⏱️

Duration

15-20 minutes

🎯

Task

Classify 15 emails as phishing or safe

🔬

Goal

Compare human and AI detection accuracy

Research Consent

This study examines how humans and AI work together to detect phishing emails. Your participation will help advance cybersecurity research.

  • ✅ All responses are anonymous
  • ✅ You can stop at any time
  • ✅ No personal data collected
  • ✅ Research data stored securely

Study Instructions

🎯 What You'll Do

You will review 15 emails and classify each as PHISHING or SAFE.

The study has three different test modes to compare different approaches.

Mode 1: Human Only

You make decisions without AI assistance

  • Review email content
  • Click Phishing or Safe
  • Rate your confidence

Mode 2: AI Only

You decide first, then see AI prediction

  • Make your decision
  • See AI prediction after
  • Option to change your mind

Mode 3: Hybrid

AI assistance shown before you decide

  • See AI prediction immediately
  • Make informed decision
  • Note if AI influenced you

🔍 Phishing Detection Tips

Suspicious Links

Check if URLs match the sender

Urgent Language

"Immediate action required" often indicates phishing

Personal Information

Legitimate companies rarely ask for passwords via email

Grammar Errors

Poor spelling and grammar are red flags

Email 1 of 15 Human Only
1
Guess 2
Unsure 3
Somewhat 4
Confident 5
Very Sure
Confidence Level: 3/5
⏱️ 0s

Post-Study Survey

Please help us improve by answering a few questions.

🎉 Study Complete!

Thank you for participating in our research

85%
Your Accuracy
92%
AI Accuracy
78%
Agreement Rate

Performance Summary

👤 View Portfolio

Research Note: Your anonymous responses have been recorded and will contribute to cybersecurity research. The AI model continues to learn from diverse human decisions.