Skip to content
PhiloCyber logo
Back to all services05 / 05

AI Security Enablement

Hands-on workshops that give engineering, product and security a repeatable way to review AI systems before release.

Scope
One intensive session of one to four hours, or weekly sessions of one to two · Format and topics chosen with you, around what your teams need
Built for
Engineering, product and security teams that need a practical baseline before they scale GenAI, agents or MCP integrations.
Plan a workshop
Animated enablement loop moving a team from baseline through practice, review and reuseBaselinePracticeReviewReuseCapability retainedTeam01Baseline02Practice03Review04Reuse

What this service does

I build the workshop around your stack and delivery process. The session uses concrete architectures, attack paths and review exercises so participants learn to spot risky agent behavior, excessive permissions, untrusted context and weak controls before release.

The value

The capability stays with the team: a shared vocabulary, a repeatable review method and artifacts people can use on the next feature. That reduces dependence on one-off reviews and brings better security questions into the design process earlier.

How it runs

One intensive session, or short recurring sessions spread across a few weeks. Format and topics are chosen with you, based on what your teams actually need. The goal is capability that remains with the team after the workshop ends.

I gave a talk on this

MCP Security for enterprise organizations, UTN 2025

Questions this work answers

  • Can the team identify AI-specific trust boundaries and abuse paths?

  • Do reviewers know what to ask about tools, permissions, RAG and memory?

  • Can the team turn a finding into a practical design change?

What you get

04
  1. D/01Workshop shaped around your architecture, tools, roles and delivery process
  2. D/02Hands-on exercises covering trust boundaries, prompt injection, tool permissions and untrusted context
  3. D/03A concise review method and checklist participants can reuse on the next AI feature
  4. D/04Written recap of decisions, capability gaps and the highest-value next actions

All services

Something already in production you are not sure about?

Tell me what you are building and what worries you about it. If it is not something I can help with, I will tell you that too.

Plan a workshop
AI Security Enablement | Ricardo N. Prieto | PhiloCyber