Argue about risks together, and put honest numbers on them.

Your team builds the assessment in one shared graph of risks, attacks and defences, and the report it produces can go to a manager or an auditor. ArgueSecure is free, open-source software you run on your own server.

The demo is public and shared. Do not enter sensitive or real data; it may be cleared at any time. About the demo.

Target · Customer portal

Risk · Customer data is leaked€250k / event

Attack · Phishing for admin credentials0.8 / year

Defence · Two-factor sign-in−80%

Transfer · Cyber insurance policy

Illustration. Expected annual loss: inherent €200,000, current €40,000.

How it works

  1. 1 · TargetWhat you are assessing: a system, a product, a process.
  2. 2 · RisksThe losses you care about, each with its cost per occurrence.
  3. 3 · AttacksClaims about how each risk happens, and how often per year.
  4. 4 · Defences & transfersCounter-claims that reduce a risk, or shift the loss to someone else.

The numbers stay honest: expected annual loss is how often an attack happens × the loss each time × the share each defence leaves, shown three ways (inherent, current, target). A missing value is flagged and left out of the total, so a gap in the data cannot look like zero risk.

It is a lightweight way to work. A team can start with a whiteboard-style brainstorm, bring in what engineers, product owners, security people and management each know, and add numbers later, because nothing needs prior quantitative data. People who cannot make the meeting add their part afterwards. The claims and counter-claims stay in the assessment as a record of how each conclusion was reached.

How it works, in detail

What teams use it for

Chart view of an example assessment while two people work in it: the avatar of the colleague who has it open is in the header, and the attack that colleague is editing shows a lock.

Running a risk workshop

Everyone opens the same assessment and sees who else is in it. One person edits a node at a time, and comments with @mentions keep the discussion next to the item it is about.

Collaboration features

Writing it up

The report shows expected loss with no defences, with what is in place and with what is planned, and ranks defences by cost benefit. Every change is in the history, and earlier versions can be compared and restored.

Reports and history

The report's executive summary, with inherent, current and target annual loss and the cost and benefit of the planned controls.
The report's Control coverage table for an ISO/IEC 27001:2022 assessment, with some Annex A controls implemented and the rest planned.

Checking against a standard

Pick a standard's template and each control becomes a defence under the threats it addresses. Mark it implemented, planned or out of scope, and the report shows what is still open. Start an assessment from the whole template and prune it, or copy only the risks and defences you need into one you already have.

Standards and templates

Features

What ArgueSecure 2.7.0 does, in short. The full list, with screenshots.

Live collaboration

Everyone sees who is in the assessment, one person edits a node at a time, and comments with @mentions stay next to the item they are about.

Arguments as a graph

An attack can serve several risks and a defence several attacks, and a shared item is edited once.

Honest quantification

Expected annual loss is shown as inherent, current and target, and a missing value is flagged and left out of the total.

Reports and exports

A report ranks defences by cost benefit, and an assessment exports to HTML, CSV, JSON, Excel or Word.

History, versions and undo

Every change is recorded, versions can be compared and restored, and structural changes can be undone.

Standards templates

Nine bundled templates, from ISO/IEC 27001 to the EU Cyber Resilience Act, turn each control into a defence to work through.

An optional local assistant

A model on your own server can suggest risks, attacks and defences, and nothing it writes is used until a person accepts it.

Two-factor and single sign-on

Accounts are created by administrators, and sign-in can use an authenticator app or your OpenID Connect provider.

Free, open source, on your own server

ArgueSecure is MIT-licensed and sends no telemetry. You run it with Docker Compose from prebuilt images on the GitHub Container Registry, on anything from a home server to a data centre, and your assessments never leave it.

Source code on GitHub

Installing it

curl -fsSLO …/v2.7.0/docker-compose.yml
curl -fsSL …/v2.7.0/.env.example -o .env
docker compose up -d
The full quick start, pinned to 2.7.0

An optional AI assistant runs on the same server; see Features.

Background

ArgueSecure grew out of research into argumentation-based security risk assessment at the University of Twente, carried out in the EU FP7 project TREsPASS (Technology-supported Risk Estimation by Predictive Assessment of Socio-technical Security).

During the project, the method was built into a desktop tool and a first online tool. Today's ArgueSecure is a complete rewrite from scratch, built on the concepts of those original tools.

Created by Dan Ionita, with Roeland Kegel, Andrei Baltuta and Roel Wieringa. About the project and how to cite it.

D. Ionita, R. Kegel, A. Baltuta and R. Wieringa, “ArgueSecure: Out-of-the-Box Security Risk Assessment,” IEEE REW 2016, pp. 74–79. doi:10.1109/REW.2016.027

The research behind ArgueSecure received funding from the European Union’s Seventh Framework Programme (FP7) under grant agreement no. 318003 (TREsPASS), and from NWO and KPN.

Look around, then run your own

The demo is a shared instance you can open now. When you want your own, the quick start takes a few minutes with Docker Compose.