Prompt Security
GenAI firewall: prompt-injection defense, DLP, and shadow-AI control.
Acquired — still operating
acquired by SentinelOne 2025
About
Prompt Security, acquired by SentinelOne in 2025, is a GenAI security platform that inspects AI traffic in real time to enforce policy on both sides of the prompt. It blocks prompt injection and jailbreaks via an inline application firewall, redacts sensitive data leaving through prompts and responses (contextual DLP), discovers shadow AI usage across browsers and desktops, and governs agentic AI — discovering agents, tools, and MCP servers and enforcing least-privilege access with an audit trail.
It is model-agnostic across OpenAI, Anthropic, Microsoft, Google, and self-hosted models. As an enterprise product now inside SentinelOne's portfolio, setup involves coordinating security and IT for endpoint/browser deployment and policy configuration.
What it can do
- Inline prompt-injection and jailbreak blocking
- Contextual DLP for prompts and responses
- Shadow-AI discovery across browsers and apps
- Agentic AI discovery and least-privilege enforcement
- Model-agnostic coverage (OpenAI, Anthropic, Microsoft, Google)
Setup
- Contact SentinelOne for Prompt Security licensing.
- Deploy the browser extension and endpoint agent for workforce coverage.
- Route AI application traffic through the inline API proxy.
- Define DLP and prompt-injection policies.
- Review shadow-AI discovery and agent audit records.
Prerequisites
SentinelOne enterprise licensing; IT coordination for endpoint deployment.
Use cases
Works with
Trust score — 6/10
Scored 6/10 (metrics tier, metrics-v1). Evidence: Acquired by SentinelOne in 2025; now its GenAI security product; Covers employee AI use, AI application security, and agentic AI (MCP discovery, least-privilege); Model-agnostic across OpenAI, Anthropic, Microsoft, Google, self-hosted; Enterprise deployment requires security/IT coordination. License: Proprietary (SentinelOne).
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 6.0/10
- Maintenance · recency, releases, issue load
- 7.0/10
- Openness · license + docs
- 4.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Acquired by SentinelOne in 2025; now its GenAI security product
- · Covers employee AI use, AI application security, and agentic AI (MCP discovery, least-privilege)
- · Model-agnostic across OpenAI, Anthropic, Microsoft, Google, self-hosted
- · Enterprise deployment requires security/IT coordination
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.