AI governance, compliance and security in the enterprise, explained without the fog.
What really happens to a prompt after you send it: three layers of retention, a court order, and the four levers that do depend on your company.
José Manuel Robles Hermoso ·
Shadow AI is not just someone pasting a contract into ChatGPT. Five routes your data takes out with nobody deciding anything, and which one to audit first.
José Manuel Robles Hermoso ·
Shared Claude conversations and Artifacts turned up in search engines. There was no breach: there was a share button and a technical control that never did what it looked like it did.
José Manuel Robles Hermoso ·
Stanford's AI Index 2026 reveals the gap between AI capabilities and governance is widening. 362 documented incidents in 2025, only 11% of companies without policies... but 59% lack the knowledge to implement them.
José Manuel Robles Hermoso ·
Shadow AI is the invisible crisis inside your enterprise. 80% of workers use unapproved AI tools and the EU AI Act starts enforcement in August 2026.
José Manuel Robles Hermoso ·
Live demonstration of how a prompt injection can compromise an enterprise AI agent in seconds. OWASP's #1 vulnerability for LLM applications.
José Manuel Robles Hermoso ·
An internal Meta AI agent posted inaccurate technical guidance without human approval, causing a two-hour data breach. Full incident analysis and lessons learned.
José Manuel Robles Hermoso ·