Security research

Senthex Research

We are a small European team building a runtime AI firewall — and we publish what we find. These are reproducible studies of how LLM and multi-agent systems actually fail in production, written for the security leaders who have to defend them. Every figure is traceable to a released dataset.