Continuous Threat Modeling: Let Developers Figure It Out

Izar Tarandach
Izar Tarandach
Sr. Principal Security Architect,
Abstract

Threat Modeling has customarily been seen as a black art,a bit of an arcane discipline that not many are privy to. And that is, basically, wrong. Everyone threat models, all the time. And they very well should!\n\nIn this talk we will look at a couple of traditional Threat Modeling methodologies, what they're good for, what they miss, and offer a new one that your developers can run with - agile and principle-based. \nAfter that we will look at a threat-modeling-with-code tool, OWASP pytm, that can be used to support continuous threat modeling by your teams, see how it helps and what it doesn't do.

00:00:00 Introduction and speaker background

00:04:40 Starting with why: the Apollo 13 story

00:10:20 What threat modeling is and is not

00:16:00 Who should participate in threat modeling

00:22:00 System modeling and threat elicitation

00:28:20 Asking what could go wrong

00:33:00 Writing useful threats and mitigations

00:38:00 Using threat models across the SDLC

00:46:00 Threat Modeling Manifesto and common failures

00:52:00 How to know threat modeling is working

00:54:40 The four questions methodology

01:00:00 STRIDE overview and exercise

01:04:00 PASTA and choosing the right methodology

01:07:30 Continuous Threat Modeling explained

01:14:00 Handbooks, checklists, automation, and AI tools