Threat modeling is the practice of working out who is likely to attack you, how they would do it and what that means for your defenses. Intelligence-driven threat modeling grounds those answers in real threat intelligence about active groups and their techniques, rather than in generic lists of everything that could go wrong.
In application security, threat modeling often means analyzing a system design with methods like STRIDE. For security operations, the question is different: given the attackers we are likely to face, are we able to detect and stop what they actually do?
Detection teams cannot cover every technique equally. A threat model tells them where to spend effort first. If the groups targeting your sector rely on stolen credentials and remote management tools, coverage for those techniques matters far more than coverage for exotic exploits nobody has used against companies like yours. It also gives SIEM modernization a clear test: does the new approach improve coverage of the threats in the model?
This is an illustrative example. Your own table comes from your threat intelligence and detection inventory.
How is this different from STRIDE?
STRIDE is a framework for finding design weaknesses in a system. Intelligence-driven threat modeling for the SOC starts from real attackers and their techniques, and asks whether you can detect them.
Who should own the threat model?
Usually the CTI team and detection engineering together, with input from the SOC and system owners.
How often should it be updated?
At least quarterly, and whenever intelligence shows a significant change in who is targeting you or how.
Anomali ThreatStream Next-Gen provides actor profiles and ATT&CK mappings that feed threat modeling, and Anomali Cyber Watch tracks the TTPs behind current campaigns each week.
Request a demo to see intelligence-driven coverage in Anomali.