An assurance case is a claim supported by a reasoned and compelling argument based on a body of evidence, offering assurance about a system, product, service or process. When it concerns safety, it is called a safety case; when it concerns security, a security case. It often takes the form of a report, sometimes with the argument shown as a tree-shaped graph or table.
A safety case is the complete package (the claims, the argument and the evidence) that shows a system is acceptably safe. The safety case is usually delivered as a report. The argument is the reasoning at the heart of the safety case. It connects the top-level safety claim to the evidence and explains why that evidence is sufficient.
CSL offers training in structured argumentation using notations such as Goal Structured Notation (GSN) and Eliminative Argumentation (EA), along with team coaching and argument reviews. CSL’s published papers and example arguments are available on the Socrates Perspectives page. Established guidance documents such as MISRA’s Guidelines for Assurance and Safety Arguments (GASA) are an additional resource supporting the development of reliable assurance cases.
Socrates supports Goal Structuring Notation (GSN), Eliminative Argumentation (EA), Claims Arguments Evidence (CAE) and Friendly Argument Notation (FAN).
Socrates helps to establish confidence in the logical soundness and completeness of an argument that demonstrates identified risks have been adequately controlled.
Yes. There is nothing built into Socrates that limits its use to just make safety arguments. It has been used by customers to make assurance arguments that address both safety and security in an integrated manner, as well as stand-alone security assurance arguments.
Start with the Socrates Product Overview video or request a demonstration to see Socrates. CSL also offers training, team coaching, and support. We respond to inquiries within one business day.
Most new users can find their way around an argument within minutes, and everyday tasks take just a few clicks. More advanced features may take a little longer to master, but teams can be productive long before they have learned everything Socrates offers. Every Socrates user has access to the Socrates Knowledge Base, an online guide with step-by-step tutorials and short how-to videos for every feature. CSL also offers training and coaching to help teams get up to speed quickly.
Users can view their work in a graphical or tabular format. Socrates can also generate reports in a variety of formats for reviews and formal assessments.
Socrates automates many routine parts of assurance case development, so teams can build arguments faster, validate their structure, and identify potential weaknesses early. Engineers remain in control of every judgment.
Socrates’ AI features are entirely optional, and the platform remains a fully functional, comprehensive solution without them. If you choose to enable AI features, you can select and configure the LLM that best meets your organization’s requirements.
Socrates can use AI to speed up routine work and to challenge your thinking: it can draft argument structures, suggest reasons for doubt, review content, and summarize parts of the argument.
Time savings grow with the size of your assurance case and how often it changes. Socrates removes the need to manually redraw diagrams after each change. Automated checks catch problems early, while reusable patterns and customizable static analysis rules help teams apply their organizational processes consistently.
Socrates is a complete, self-contained assurance case management solution with no third-party subscriptions or products required.
Yes. Socrates supports a range of sharing options, including Word reports, image exports, JSON and CSV exports, and live read-only access. All of these can be filtered to present only the information relevant to a particular audience.
Yes. Legacy arguments can be imported into Socrates, provided your current tool can export the underlying argument structure and not just a final diagram. With the Socrates Python REST API, migrating existing assurance case content is straightforward, and CSL can assist throughout the import process.
Socrates has a number of mechanisms that allow users to manage very large, complex arguments. Users can divide an argument into sub-arguments that are opened and edited separately, filter the view to focus on specific concerns, and save repeated structures as reusable patterns, ensuring those structures remain consistent.
Any type of file can be linked to your assurance case; there are no file constraints. Socrates links to your evidence wherever it lives (test reports, design reviews, SharePoint documents, web pages or records in other tools) rather than storing copies. Numerical data, such as test data or field results, can also be attached as live indicators.
Yes. When several people work on the same argument, each person’s changes show up for everyone else automatically, without refreshing. Updated nodes are highlighted, and notifications are sent through the application, by email, or both. You are notified when someone edits or deletes part of the argument, links evidence to it, comments on it, replies to you or mentions you. Each user can choose which notifications they receive, allowing reviewers to stay informed at their preferred level of detail.
Socrates has four user roles:
More information can be found on our website.
Multiple arguments can be created within an instance of Socrates. By default, arguments are private with access restricted to specific users by invitation. However, any argument can grant public access to all users if desired.
Yes. Socrates integrates with your toolchain through the Socrates Python REST API. Testing, simulation, or monitoring data can feed live indicators in the argument. Arguments can be imported and exported in formats such as JSON, CSV, and DOCX.
Socrates can be deployed either on-premises or in the cloud. Most customers choose an on-premises deployment, where Socrates is installed on their organization’s infrastructure behind their corporate firewall. Critical Systems Labs also offers cloud-hosted deployments for organizations that prefer a managed hosting option.
Most organizations deploy Socrates on their own infrastructure behind their firewall, retaining full control over security, access management, and data. Being web-based simply means users access the system through a browser, which enables collaboration while centralizing administration and access control.
Yes. Modern standards and regulations increasingly require a structured safety case, which is exactly what Socrates is built for. Socrates is not tied to any one standard, making it well suited for frameworks such as ISO/PAS 8800, UL 4600, UN R185 and other industry-specific standards and frameworks. For a practical example, see From Standard to System: ISO/PAS 8800 Assurance in Practice with Socrates, which demonstrates how Socrates can be used to support compliance with ISO/PAS 8800.
Socrates has been used in automotive, aerospace, nuclear, defense, energy, maritime, cybersecurity and scientific research, including at CERN for the Large Hadron Collider. Socrates is most valuable for complex or novel systems, such as those using AI, when many contributors supply evidence, and when the argument must stay current over a long product life.
Socrates is not designed specifically for STPA, but it complements STPA well. Results from STPA, such as Unsafe Control Actions (UCAs), can be incorporated directly into safety arguments and traced to the evidence and mitigations used to address them. Depending on the argumentation approach, the concerns identified during STPA may be represented as defeaters, counterclaims, or counter evidence.
Some standards, such as the rail standard EN 50129, define what a safety case must contain rather than asking for an argument. Socrates can follow that prescribed structure while adding traceability, version control, and collaboration. For novel technology such as machine learning, where checklists fall short, an explicit argument shows assessors that the hard questions have been asked.
A “live safety case” is a safety case that is connected to real-world data. Socrates attaches Key or Safety Performance Indicators (KPIs/SPIs), such as test results or field data, to the claims they support and flags any that cross a threshold. The safety case becomes a dashboard that stays current, rather than a document that goes out of date.
Yes. Instead of keeping a separate, duplicated assurance case for each product variant, Socrates lets you keep one common argument and mark where each variant differs in claims, evidence or monitored performance data. You can then instantly see a focused view, showing only what applies to each variant. Shared parts are maintained once, so changes are automatically reflected across all variants in the product line. More information can be found here.
