TrustLens

Inspiration

Cybersecurity warnings are often written for people who already understand cybersecurity.

A student receiving a suspicious internship email, a fake university announcement, or a message asking them to “verify” their account may see a warning saying “Potential phishing detected” — but that does not necessarily tell them why the message is dangerous or what they should do next.

This led us to ask:

What if cybersecurity tools didn't just detect threats, but taught people how to recognize them?

We wanted to create a solution that makes cybersecurity more understandable and actionable, particularly for students and other users who may not have a technical background.

That idea became TrustLens.

What TrustLens Does

TrustLens is a conceptual cybersecurity assistant designed to help users identify phishing, scams, and social-engineering attacks while explaining the reasoning behind its assessment.

Instead of simply presenting a technical warning, TrustLens would translate security signals into clear, human-readable guidance.

For example, rather than:

THREAT LEVEL: HIGH

TrustLens could explain:

• The sender's domain does not match the organization being impersonated.
• The message creates urgency and asks the user to act immediately.
• The included link redirects to a suspicious domain.
• The message requests sensitive information.

It would then provide an actionable recommendation, such as avoiding the link and verifying the request through an official channel.

The goal is not to make users dependent on another security tool. The goal is to help them become better at recognizing threats themselves.

How We Developed the Idea

We started by looking at common forms of social engineering and asking why existing security warnings can be difficult for non-technical users to interpret.

From there, we mapped the problem into three stages:

1. Detect — identify potentially suspicious characteristics of a message or digital interaction.
2. Explain — translate those characteristics into language a non-expert can understand.
3. Act — provide a clear, practical recommendation for what the user should do next.

We then designed the concept around a simple user experience rather than a complex security dashboard. This helped us focus on accessibility and education alongside technical capability.

Because this is an ideathon, our submission focuses on the proposed system architecture, user experience, implementation plan, and potential impact rather than claiming a fully deployed product.

Technology

The proposed version of TrustLens could combine several cybersecurity and machine-learning techniques.

Potential analysis components include:

• URL and domain analysis
• Phishing-indicator detection
• Sender and message metadata analysis
• Natural-language analysis of suspicious messages
• Machine-learning-based risk classification
• Explainable AI techniques to identify the factors contributing to a risk assessment

The output would be converted into a simple risk explanation rather than exposing users to raw technical data.

A future implementation could also incorporate feedback from users to improve detection and identify emerging social-engineering patterns.

What We Learned

One of the biggest lessons from developing TrustLens was that effective cybersecurity is not only about detecting threats.

A technically accurate warning can still fail if the person receiving it does not understand what it means.

This made us think about cybersecurity as both a technical and human problem. Detection, explainability, accessibility, and user behavior all have to work together.

We also learned how important it is to define a problem before choosing a technology. Instead of starting with “How can we use AI?”, we started with “Why do people still fall for threats even when security tools exist?” That shift helped us develop a more focused solution.

Challenges

One of our biggest challenges was balancing security accuracy with simplicity.

A cybersecurity system needs to consider many signals when assessing a potential threat, but presenting every technical detail to a non-technical user could make the system overwhelming.

Another challenge is avoiding false positives. Incorrectly labeling a legitimate message as malicious could cause users to lose trust in the system.

Privacy is also an important consideration. A real implementation would need to minimize the amount of sensitive user data collected, process information securely, and clearly communicate how user data is handled.

These challenges shaped our proposed design and reinforced the importance of building TrustLens as a decision-support and education tool, rather than an infallible security authority.

Future Potential

TrustLens could eventually expand beyond email phishing.

Potential applications include:

• Fake internship and job offers
• Fraudulent university communications
• Suspicious social-media messages
• Malicious QR codes
• Account-verification scams
• Impersonation attempts
• Other forms of social engineering

The long-term vision is to make cybersecurity guidance as understandable as possible — giving people not only a warning, but the knowledge needed to make safer decisions.

TrustLens is built around a simple idea:

Don't just tell people that something is dangerous. Help them understand why.

Built With

• ai
• cybersecurity
• design
• detection
• engineering
• experience
• human-computer
• language
• learning
• machine
• natural
• phishing
• processing
• social
• threat
• user