Field noteIn preparation
The State of AI Security in Morocco
What Moroccan organizations need to secure as generative AI moves from pilots into operations.
Link Security Lab
Briefs for people who deploy AI.
A brief is published when it is ready to defend. The list is the work in preparation.
Research for people who deploy AI.
Attack techniqueIn preparation
How Prompt Injection Attacks Work
How untrusted text becomes an instruction, and what that means for any system that reads it.
ArchitectureIn preparation
Securing RAG Systems
Where retrieval expands the attack surface: documents, permissions and the answers a model will repeat.
ArchitectureIn preparation
Why AI Agents Change the Cybersecurity Model
Agents do not only answer. They call tools. That moves AI from a feature to a privileged user.
Attack techniqueIn preparation
Shadow AI: The New Enterprise Attack Surface
Unapproved copilots and personal accounts are already inside organizations. They need an owner.
Field noteIn preparation
AI Security for Moroccan Organizations
A practical frame for banks, hospitals, public bodies and companies adopting AI under Law 09-08.
GovernanceIn preparation
Digital Sovereignty and AI Security
Sovereignty is control of data, models, access and the evidence around them.
Contact
Start with a conversation.
No system access is required to start. This form is delivered to nourkabbouri@protonmail.com. Nour Kabbouri reads it. The founders lead the work that follows.