An introduction into the Automotive Threat Matrix

The Automotive Threat Matrix (ATM) is a free, public catalog of the tactics and techniques attackers use against vehicles, run by the Automotive Information Sharing and Analysis Center (Auto-ISAC). It works like MITRE ATT&CK, but only for motor vehicles, and every technique has its own ID so different teams can name the same attack step the same way. Security teams use it to find attack paths in a threat analysis and risk assessment (TARA), plan tests and share threat intelligence.

What is the Automotive Threat Matrix?

Auto-ISAC, the industry group where automakers and suppliers share information about vehicle cyber threats, launched the ATM publicly in March 2024 at atm.automotiveisac.com. Auto-ISAC's 2025 white paper calls it "a globally accessible knowledge base offering a common language and taxonomy of adversary tactics," curated by industry experts.

The ATM is modeled after MITRE ATT&CK, the public knowledge base of attacker behavior in IT networks, and keeps its structure while focusing strictly on motor vehicles. It is a living tool: an ATM User Group, open to people beyond Auto-ISAC members, reviews proposed changes before they are adopted.

How the matrix is organized: tactics, techniques and IDs

Each column is a tactic, the goal an attacker has at that step. Auto-ISAC's March 2026 briefing counts 14 tactics. When we first wrote this post, the matrix showed these 13, so check the live matrix for the current set:

• Manipulate environment
• Initial access
• Execution
• Persistence
• Privilege escalation
• Defense evasion
• Credential access
• Discovery
• Lateral movement
• Collection
• Command and control
• Exfiltration
• Affect vehicle function

Under each tactic sit techniques, the "how." Here is what you see inside the matrix (pun intended) when you select Discovery:

Open one technique, such as Network service scanning, and you get a description, examples and an ID:

Technique IDs take the form ATM-T followed by four digits. In our screenshot, Network service scanning is ATM-T0044. When two teams at two companies both write ATM-T0044, they know they mean the same thing.

How to use the ATM in TARA, testing and intelligence sharing

Our thanks go to Karl Leboeuf for his presentation on the ATM at the 1st Annual European Cybersecurity Summit, and to Josh Poster for his talk on it. The uses below draw on their presentations and on Auto-ISAC's February 2025 ATM white paper, which also covers cybersecurity testing, vulnerability management, incident response and compliance reporting.

TARA and ISO/SAE 21434

ISO/SAE 21434 is the international standard for cybersecurity engineering of road vehicles. Its clause 15 defines TARA as seven methods:

• Asset identification: the assets, the cybersecurity properties you care about, and the damage scenarios if they are compromised
• Threat scenario identification: how each property could be compromised
• Impact rating of each damage scenario
• Attack path analysis: the steps that would realize each threat scenario
• Attack feasibility rating for every attack path
• Risk value determination
• Risk treatment decision

The standard does not tell you how to find threats or attack paths, and it does not list threats you must consider. A note in clause 15.4.2 says threat scenarios can come from group discussion or systematic approaches, such as misuse cases or frameworks like EVITA, TVRA, PASTA and STRIDE. In our experience STRIDE is the most prescriptive of these, but it gives you categories (spoofing, tampering and so on), not the specific moves an attacker would make against a telematics unit or a Controller Area Network (CAN) bus.

That is the gap the ATM fills. It gives you an observed catalog to draw threat scenarios from, and attack paths can be built as chains of ATM techniques. This feeds straight into UN Regulation No. 155, the vehicle cybersecurity type-approval rule, which requires an exhaustive risk assessment for each vehicle type. Once the paths are rated, the next step is deciding what to do about each risk, which we cover in Navigating the Risk Treatment Decision in TARA.

Security engineers

Because the techniques are verified attack methods curated by industry experts, rather than a brainstorm, the matrix works as a prompt list when you elicit attack paths, so an important path doesn't get left out. It helps less experienced engineers most: it nudges them to consider a broader set of threats than they would come up with alone. That pairs well with bringing design engineers into the TARA early, as we describe in Involving Engineers Early in the TARA Process.

The talks also pointed to where this is heading. As the ATM matures, most techniques should gain matching mitigations, and future work could include an automotive counterpart to MITRE D3FEND, MITRE's public catalog of defensive techniques.

Intelligence sharing and incident response

When a new attack shows up, the matrix gives analysts a common way to describe it:

• Identify the known tactics and techniques used in the attack
• Spot new tactics and techniques not seen before
• Understand why and how the adversary may be taking each step
• Use common definitions so everyone knows what everyone else means
• Write reports faster by reusing the ATM's dataset, references and resources

This is the same shared language a vehicle security operations center and a product security incident response team need when they hand work to each other. More on that in Bridging the Gap: VSOC and PSIRT in Automotive Cybersecurity.

How the ATM is updated and how to contribute

Auto-ISAC's white paper is clear that neither threat models nor the ATM are single-use tools, and the matrix changes as new attack paths and research appear. Anyone can propose a change by selecting a technique and adding their insight:

As presented in the 2024 talks, submissions are reviewed monthly by the ATM working group, a review panel of subject matter experts approves recommendations each quarter, and additions go out to the community through the mailing list and release notes on the website. Auto-ISAC's 2026 material sums it up the same way: proposed changes get a group review before adoption.

Where to start with the Automotive Threat Matrix

The easiest first step is to use the ATM on work you've already done. Take one finished TARA, map each attack path to ATM technique IDs, then scan each tactic for techniques your team never considered. The gaps you find are your next threat scenarios.

We've run nearly 400 TARAs across automotive, agriculture and robotics, from full vehicles down to single chips. If you'd like a second pair of eyes on how your team finds attack paths, tell us about the TARA you're working on and we'll talk it through.