Open Source

Research / Open Source

Make the work open.
Make it useful.

Trustworthy AI improves when people can inspect methods, understand assumptions and examine the evidence behind a result. We share selected tools and knowledge that help research become useful in practice.

Explore TrustInspect Discuss a contribution
Conceptual research campus with connected glass pavilions.
Our open-source vision

Help others examine, repeat and improve the work.

  1. Inspect the method

    Document what an approach does, the assumptions it relies on and the conditions in which its results are useful.

  2. Repeat the work

    Connect tests with their conditions, observations and evidence so teams can repeat relevant work and compare findings.

  3. Improve through use

    Use feedback, practical examples and contributions to identify gaps and strengthen the methods over time.

Featured initiative

TrustInspect

Stylised presentation of TrustInspect’s report and terminal interfaces.
Open-source release planned

From AI behaviour to reviewable evidence.

TrustInspect connects testing for LLM, RAG and agentic applications with prompts, responses and findings that teams can investigate.

The public release is being prepared. Repository access, licence and contribution guidance will accompany the release.

Explore TrustInspect Register interest
Future contributions

An open programme shaped by the research.

Areas of interest include assessment support, testing methods and evidence collection across the AI lifecycle. We will introduce further tools and materials here when they are ready to share.

Each release needs a clear starting point.

Its purpose, intended users, available functionality, limitations, licence and contribution process should be understandable before a team adopts it.

Community work

Contribute to shared foundations.

OWASP AI Maturity Assessment

Community work on the organisational practices needed to develop and use AI responsibly.

OWASP AI Testing Guide

Community work on practical methods for testing AI applications and interpreting findings.

Matteo Meucci is Creator and Contributor of both projects. These are OWASP community initiatives, distinct from Synapsed’s frameworks and tools.

Open knowledge. Protected innovation.

We share selected tools, methods and knowledge while protecting innovations that require sustained research investment. The scope and licence of each release explain how it can be used.

Discuss collaboration Explore our research