# GenAI Security Finder

> A finder for controlling employee and AI agent use of generative AI: answer seven questions to see which approach fits (AI controls in a SASE or SSE cloud, a standalone AI security layer, or a firewall or self-hosted AI gateway) and which of six vendors' published facts match: Cato Networks, Cisco, Fortinet, Netskope, Palo Alto Networks and Zscaler. No scores or rankings. Vendor pages read 7 October 2026.

## Key pages

- [Home](/): Where a GenAI control can sit, the quiz questions and the vendors covered.
- [GenAI security quiz](/genai-security-quiz): Seven questions that point to a control approach and the vendors whose published facts match.
- [Finder rules](/finder-rules): The published rules behind the quiz, the evidence and its limits.
- [Control approaches](/control-approaches): The three places a GenAI control can sit and which vendors describe each.
- [AI controls in a SASE or SSE cloud](/control-approaches/sase-sse-ai-controls): What controls in the security cloud see and where they stop.
- [Standalone AI security layer](/control-approaches/standalone-ai-security): AI gateways and runtime guardrails for homegrown AI apps and agents.
- [Firewall or self-hosted AI gateway](/control-approaches/firewall-self-hosted): Keeping AI traffic, prompts and logs on infrastructure you run.
- [Vendor profiles](/genai-profiles): Six vendors' stated AI controls, with sources, strengths and limits, plus every vendor pair.
- [Facts side by side](/facts-side-by-side): Two to six vendors' published facts in one table.
- [Explainers](/genai-explainers): Eight short explainers in three tracks: concepts, buying and running it.
- [AI usage policy checklist](/ai-usage-policy-checklist): The decisions to make before you buy a control.
- [Glossary](/genai-security-glossary): Plain definitions of GenAI security terms, from A2A to ZTNA.
- [Questions](/genai-security-questions): Short answers on shadow AI, agents, the vendors covered and this site.
- [Cato Networks and Cisco](/cato-networks-and-cisco-genai-controls): An example pair page; all fifteen vendor pairs are linked from the vendor profiles page.
- [Piloting GenAI controls](/genai-explainers/piloting-genai-controls): Scope, success measures and tests for a pilot.
- [Inline Notes](/inline-notes): Dated analysis of what the six vendors state about agents, MCP, FedRAMP, consoles and discovery.
- [Contact](/contact): Questions, corrections to vendor information, privacy requests and copyright notices.
- [Privacy Policy](/privacy-policy): What personal data the site handles, and your rights.
- [Cookie Policy](/cookie-policy): No cookies; one functional storage key for the checklist.
- [Terms of Use](/terms-of-use): Terms for using the site.
- [Copyright and DMCA policy](/dmca): How to send a copyright notice or counter-notification.

## Vendor profiles

- [Cato Networks profile](/genai-profiles/cato-networks): What Cato's pages state about AI controls.
- [Cisco profile](/genai-profiles/cisco): What Cisco's pages state about AI controls.
- [Fortinet profile](/genai-profiles/fortinet): What Fortinet's pages state about AI controls.
- [Netskope profile](/genai-profiles/netskope): What Netskope's pages state about AI controls.
- [Palo Alto Networks profile](/genai-profiles/palo-alto-networks): What Palo Alto Networks' pages state about AI controls.
- [Zscaler profile](/genai-profiles/zscaler): What Zscaler's pages state about AI controls.

## Explainers

- [What is shadow AI](/genai-explainers/shadow-ai-explained): Unapproved AI use, why it happens and how teams discover it.
- [Prompt DLP](/genai-explainers/prompt-dlp-explained): How sensitive data is caught in prompts, and what redaction adds.
- [MCP and AI agent risk](/genai-explainers/mcp-agent-risk-explained): Why agents and MCP servers need controls that chat tools did not.
- [Inline versus out-of-band](/genai-explainers/inline-vs-out-of-band): Why the position of a control matters.
- [Questions to ask a GenAI security vendor](/genai-explainers/questions-for-genai-security-vendors): A short list that separates what a product does from what its page says.
- [Reading vendor AI claims](/genai-explainers/reading-vendor-ai-claims): Counts, superlatives and analyst badges, and what each tells you.
- [Watch, coach, redact, block](/genai-explainers/watch-coach-redact-block): Choosing the first response.
- [Piloting GenAI controls](/genai-explainers/piloting-genai-controls): Scope, success measures and tests for a pilot.

## Inline Notes

- [AI agents and MCP](/inline-notes/agents-and-mcp-what-six-vendors-state): Which vendors name MCP, which describe agent identity, and what it means for a buyer.
- [FedRAMP and GenAI controls](/inline-notes/fedramp-and-genai-controls): What the vendors' trust pages say, and why AI features need their own scope question.
- [Coding agents at work](/inline-notes/coding-agents-what-security-vendors-name): Which vendors name coding agents, and what to ask before a pilot.
- [One console or two](/inline-notes/one-console-or-two-genai-controls): How vendors package employee AI controls and AI app security.
- [SASE analyst placements and AI security](/inline-notes/analyst-placements-and-ai-security): What 2026 placements tell an AI security buyer, and what they do not.
- [Shadow AI discovery](/inline-notes/shadow-ai-discovery-six-vendors): How the vendors describe finding unapproved AI, and what to ask about coverage.

Vendor pages read 7 October 2026. No scores or rankings.