Looking at your
security posture
through a new frame
in the age of agentic AI
We help organizations rethink how they operate in a world of autonomous AI. From shaping a clear AI-native vision to building the first AI agents, we focus on making autonomy safe by design. Our collaborative purple teaming approach fosters a safe learning environment that prevents vulnerable AI systems.
About us
We have been passionately developing secure AI applications for the last 20 years.
0+
Years in AI
0+
Companies advised
0+
Industries
0+
Countries
Getting AI security right is already crucial but I believe the stakes will get significantly higher as we embed AI more deeply into our lives and organizations. Staying up to date in this field requires continuous learning and what can be more rewarding than that?
Koen Bonenkamp
AI Security Engineer at Refraim
Services
Transparent and focused on knowledge sharing
Purple teaming & security assessments
We work closely with your AI engineers to make sure autonomous AI is secure by design. We focus on knowledge sharing to set your team up for long term success.
Awareness workshops
We organize workshops to raise awareness about AI security and how the probabilistic nature of AI models changes the defensive game.
Tooling
We develop AI security tooling that makes it faster and easier to secure your applications. We are also active contributors to the open source AI security community.
