Explainers · 3 tracks
GenAI security explainers: the ideas behind the controls
In short
Eight short explainers in three tracks. Concepts covers shadow AI, prompt DLP, MCP and agent risk, and inline versus out-of-band enforcement. Buying covers the questions to ask vendors and how to read their claims. Running it covers choosing a first response and running a pilot.
Concepts
01
What is shadow AI, and how do security teams find it?Unapproved AI use, why it happens and the three ways teams discover it.
3 min read
02
Prompt DLP: how sensitive data is caught before it reaches a modelHow data loss prevention works on prompts and responses, and what redaction adds.
3 min read
03
MCP and AI agent risk: what changes when AI can actWhy agents and MCP servers need controls that chat tools did not.
3 min read
04
Inline versus out-of-band AI enforcement: why the position mattersWhat it means for a control to sit in the path, and when watching from the side is enough.
2 min read
Buying
05
Twelve questions to ask a GenAI security vendorA short list that separates what a product does from what its page says.
2 min read
06
How to read vendor AI security claimsCounts, superlatives and analyst badges, and what each one tells you.
2 min read
Running it
07
Watch, coach, redact, block: choosing the first responseFour responses to a risky AI interaction, and how to roll them out.
2 min read
08
How to run a GenAI security control pilotScope, success measures and the tests that show whether a control fits.
2 min read