Threat modeling for regulated industries

An ISO/SAE 21434 TARA, an FDA premarket threat model, a HIPAA Security Risk Analysis and a SOC 2 risk assessment all start the same way: map the system, find what could go wrong, rate it, decide what to do. ThreatTree does that work once and shapes it into the evidence your industry expects.

  • Seven industry guides
  • Industry Modules on Pro
  • Free plan, no credit card

What is threat modeling for regulated industries?

It is threat modeling that produces the specific evidence an industry's rules ask for, such as a TARA under ISO/SAE 21434, an FDA premarket threat model, a HIPAA security risk analysis or an IEC 62443 risk assessment. The method is the same everywhere: map the system, find and score threats, treat them. What changes is the vocabulary, the standards each mitigation maps to and the format of the evidence.

Industries

Choose your industry

Each guide walks through what your regulator or assessor expects, step by step, and shows where each part lives in ThreatTree.

How it fits

Model once, answer every assessor

The vocabulary changes by industry; the discipline doesn't. One ThreatTree forest becomes the work product each framework asks for, so nothing is rebuilt by hand for the next audit.

One ThreatTree forest

  1. 1
    Data flow diagramComponents, data flows and trust boundaries: your scope.
  2. 2
    Attack treesHow each threat could happen, tagged with STRIDE, ATT&CK and more.
  3. 3
    Risk registerLikelihood × impact, treatment, owner and mapped controls.
In the product

See it in your industry

Real screens with Industry Modules switched on, not mockups.

Attack feasibility on every TARA path

With the Automotive module on, each attack step gets the attack-potential factors ISO/SAE 21434 uses (elapsed time, expertise, knowledge of the item, window of opportunity and equipment) and ThreatTree rates its feasibility. Goals carry a Damage Scenario and a CAL.

  • Damage Scenario and CAL on the attacker goal
  • Feasibility rated from five attack-potential factors
  • STRIDE tags and treatment plans alongside
Read the automotive guide →
The attack tree editor with the Automotive module on: the goal ‘Unlock and start a vehicle without the owner’ splits through an OR gate into a relay attack on keyless entry, abuse of the telematics API and a malicious OTA update. The relay attack is selected, and its panel shows Attack Feasibility factors (elapsed time up to 1 week, Proficient expertise, Public knowledge, Easy window, Specialized equipment) rated High, plus a Mitigate treatment: UWB distance bounding on the key fob.

Zones, conduits and a Security Level Target

The Critical Infrastructure module adds IEC 62443 zone boundaries with a Security Level Target on each. Draw conduits between zones through the firewalls that enforce them, and every threat is anchored to the zone it crosses.

  • Zone boundaries with SL-T from SL0 to SL4
  • Conduits drawn as labelled data flows
  • Modelled offline: nothing connects to your OT network
Read the critical infrastructure guide →
A water treatment plant modelled as three IEC 62443 zones: Enterprise (SL-T SL1) with ERP and a plant data mirror, Supervisory SCADA (SL-T SL2) with an HMI and historian, and Control (SL-T SL3) with a dosing PLC and pump station RTU. Firewalls sit between the zones, with conduits labelled OPC UA, Modbus/TCP and one-way replication. The control zone is selected, showing Kind: Zone (IEC 62443) and Security Level Target SL3.

LLMs, agents and tools as first-class elements

The AI & ML module adds element types for LLM endpoints, agents, tools, RAG retrievers and vector stores, a Context Window boundary for everything that ends up in the prompt, and AI data classifications such as PII in prompts.

  • AI element types with AI-specific suggested threats
  • Context Window boundary marks prompt privilege
  • OWASP LLM and Agentic Top 10 and MITRE ATLAS tags
Read the AI & ML guide →
A customer support assistant: a Customer sends chat messages to a Support agent inside a dashed purple Context Window boundary labelled ‘High privilege: prompt context’, together with a Hosted LLM and a Knowledge retriever. Outside it, a Refund tool writes to an Orders DB and a vector store holds help-centre content. The Hosted LLM is selected, with data classification PII in Prompts.

One ranked register for every assessor

Whatever the industry, every scored step lands in one register, ranked by likelihood × impact, with MITRE ATT&CK techniques, treatment plans and a named risk owner. Export it as a PDF on any plan, or the full board-ready report on Pro.

  • Severity counts, filters by tree, risk, owner and status
  • Treatment and owner on every row
  • PDF export on every plan, including Free
See it applied to PCI DSS →
The ThreatTree risk register for a payments platform: severity counts of 1 Critical, 6 High, 0 Medium and 1 Low, filters by tree, risk, owner and status, and ranked rows such as credential stuffing against merchant login (Critical 16) with MITRE technique T1110.004, STRIDE Spoofing and owner Head of Merchant Ops.
Industry Modules

Fields and controls in your industry's language

Every guide works on the core editors, on any plan. On Pro and Enterprise, an organization owner can switch on one or more Industry Modules to add the fields, classifications and controls each standard uses.

IndustryStandardsWhat the module adds
AutomotiveISO/SAE 21434, UN R155Damage Scenario, Cybersecurity Assurance Level (CAL) and Attack Feasibility fields on attack tree nodes, plus vehicle element types (ECU, central gateway, TCU, infotainment, in-vehicle bus, OBD-II port, OTA server, V2X, digital key, fleet backend) with suggested threats mapped to UN R155 Annex 5
Medical DevicesFDA premarket cybersecurityHarm Severity and Harm Scope fields, including multi-patient harm, plus device element types (firmware, BLE link, companion app, device cloud, clinician programmer, therapy actuator, sensor, update service) with suggested threats mapped to the FDA guidance's security control categories
FintechPCI DSS v4.0, DORA, PCI Secure SLCCardholder Data (PAN) data classification, PCI DSS and DORA pinned in the standards picker, release-version tags on forest snapshots, plus payment element types (token vault, HSM, payment switch, ledger, fraud engine, open-banking API, KYC provider, payment rail) with suggested threats mapped to PCI DSS and DORA
HealthcareHIPAA Security RuleePHI data classification and HIPAA Security Rule controls, plus healthcare element types (EHR, FHIR API server, HL7 engine, PACS, patient portal, HIE, device gateway, e-prescribing) with suggested threats mapped to HIPAA
Critical InfrastructureIEC 62443, NERC CIPZone trust boundaries with a Security Level Target (SL-T), Purdue levels, plus OT element types (PLC, RTU, HMI, SCADA server, historian, engineering workstation, safety system, field device, industrial firewall, data diode, jump host) with suggested threats mapped to IEC 62443 and ATT&CK for ICS
AI & ML SystemsOWASP LLM & Agentic Top 10, MITRE ATLAS, NIST AI RMFAI element types (LLM endpoint, vector store, RAG retriever, agent, tool, MCP server, guardrail, model registry, embedding model, training and evaluation pipelines, dataset), a Context Window boundary, AI data classifications, and an AI threat coverage section in the PDF report
SaaS & StartupsSOC 2, DORASOC 2 Trust Services Criteria controls for mapping mitigations, DORA for ICT providers serving EU financial entities, plus tenancy element types (tenant isolation layer, support console, webhook delivery, SSO / SCIM connector, audit log store, feature flags) with suggested threats mapped to SOC 2

All industries can also map mitigations to ISO 27001:2022, NIST SP 800-53, NIST CSF 2.0, CIS Controls v8, OWASP ASVS, PCI DSS, SOC 2 and NIS2 on Pro. Compare plans.

What this is, and isn't: ThreatTree is a threat modeling tool, not an auditor, a notified body or a regulatory consultant. It helps you produce the analysis and evidence these frameworks ask for; whether that satisfies a specific audit or submission is your assessor's decision.

FAQ

Industry threat modeling, answered

What security, compliance and engineering leads ask before adopting ThreatTree in a regulated environment.

Still have questions?

We usually reply within a day.

Get in touch
Does ThreatTree make us compliant?

No tool makes you compliant on its own. ThreatTree helps you produce the threat identification, risk assessment and treatment evidence that ISO/SAE 21434, FDA premarket guidance, PCI DSS, HIPAA, IEC 62443 and SOC 2 ask for. Whether it satisfies a particular audit or submission is your assessor's call.

Which plans include Industry Modules?

Industry Modules are included on Pro ($29 per user per month) and Enterprise. An organization owner switches them on in Settings, under Membership & Billing, and they apply to everyone in the organization.

Can we enable more than one Industry Module?

Yes. Modules are independent toggles, so a team building a connected medical device with a payments flow can enable Medical Devices and Fintech together, and both sets of fields and controls appear in the editors.

Do I need an Industry Module to follow the industry guides?

No. Every guide is built on data flow diagrams, attack trees and the risk register, which are on every plan including Free. Modules add the industry-specific fields, classifications and controls on top, such as Damage Scenario and CAL for automotive or the ePHI classification for healthcare.

Does ThreatTree connect to our production systems or OT network?

No. ThreatTree is a modeling tool. You describe the system; nothing scans or connects to your production, clinical or operational technology networks. Enterprise customers can optionally import a first-draft diagram from Terraform, CloudFormation or OpenAPI files they upload.

How is sensitive design information protected?

Any forest can be locked with client-side AES-256-GCM encryption on every plan. When you lock it, your browser encrypts its names, labels and properties, backups included, and replaces the readable copies on the server, which never sees the key. See the security page.

Get new threat-modeling guides by email as they're published — no more than one a week.

Build your first industry threat model

Start free with the core editors, and switch on your Industry Module on Pro when you're ready.