top of page

Threat Modelling Done Right

Updated: Sep 22, 2023

Threat Modelling Done Right

In the multifaceted realm of cybersecurity, envisioning potential threats holds as much significance as crafting defensive strategies. Threat modelling, a process oriented towards identifying, quantifying, and addressing potential security risks, is pivotal. When aptly incorporated into pentesting, threat modelling amplifies the depth and range of security assessments. Here's a guide enriched with examples to embed threat modelling effectively into penetration testing:

1. Grasp the Essence of Threat Modelling

Consider a bank planning to install a new ATM system. Before doing so, they'd want to predict all potential robbery techniques. Similarly, threat modelling anticipates security challenges, ensuring a comprehensive penetration test.

2. Begin with an Overview:

Define the System: If you're modelling for a web-based retail application, map out user interfaces, databases, payment gateways, and inventory management systems.

Identify Entry and Exit Points: For our retail app, entry points might include login pages, product upload interfaces, or API endpoints, while exit points might be payment confirmations or email notifications.

3. Recognize Valuable Assets

For a healthcare portal, assets could range from patient health records to proprietary medical algorithms. Pinpointing these assets elucidates the potential targets for attackers.

4. Foresee Potential Threat Actors

A video game might attract cheaters looking to exploit game mechanics, while a financial application might be targeted by seasoned cybercriminals seeking monetary gains.

5. Enumerate Threats:

Using STRIDE for our earlier retail app example:

- Spoofing: Can someone fake a user identity to gain unauthorized access?

- Tampering: Can product prices or descriptions be maliciously altered?

- Repudiation: Can users deny actions, like orders or reviews, later?

- Information Disclosure: Are customer addresses or payment details at risk?

- Denial of Service: Can someone overload the system, rendering it unusable for genuine users?

- Elevation of Privilege: Can a regular user gain admin rights?

6. Prioritize Threats:

For an airline reservation system, an attack scenario where flight routes can be altered would be a high-impact and hopefully low-likelihood threat, requiring immediate attention.

7. Craft Realistic Test Scenarios:

For a university portal, pentesters might simulate scenarios where a student tries to alter grades or access another student's transcript based on the threats identified.

8. Iteratively Update the Threat Model:

If our retail app introduced a new chat feature, the threat model would need to be updated to address new threats like spamming or phishing via chat.

9. Develop Feedback Channels:

Post-pentest of a smart home system, if testers manage to compromise a smart fridge in a way not previously modelled, this new vector should be incorporated into the threat model.

10. Foster Collaboration and Awareness:

For instance, developers of a car's infotainment system may offer insights into vulnerabilities that security teams haven't considered. Collaboration ensures a holistic threat model.

Threat modelling, when enriched with real-world examples, becomes a tangible and relatable exercise. By vividly understanding risks and bolstered by concrete scenarios, organizations can preemptively counteract adversarial tactics. In cybersecurity, a proactive stance fortified with experiential knowledge is the best defense.


Related articles

Do you know all risks in your application?

Get a free threat modeling from our experts!

Got it! We'll process your request and get back to you.

bottom of page