Guardrails for the generative era

LLM Application Security

11 companies tracked by our intelligence team

Market Overview

LLM Application Security addresses the unique vulnerabilities introduced when large language models are integrated into production applications. This category encompasses prompt firewalls, output guardrails, content filtering, and runtime protection for any application that incorporates LLM capabilities — from customer-facing chatbots to internal copilots and autonomous agents.

The core challenge is straightforward but technically demanding: LLMs are fundamentally different from traditional software. They accept natural language input, generate non-deterministic output, and can be manipulated through carefully crafted prompts in ways that bypass conventional security controls. The OWASP Top 10 for LLM Applications has codified these risks — prompt injection, insecure output handling, training data poisoning, and excessive agency among them — providing a framework that this category's vendors are building against.

Companies like Prompt Security (acquired by SentinelOne), CalypsoAI, and Arthur AI's guardrails platform provide the runtime security layer that sits between users and LLM applications, filtering both inputs and outputs in real time. Lasso Security focuses specifically on securing enterprise GenAI interactions, while WhyLabs and Galileo provide the evaluation and monitoring infrastructure that detects when LLM applications begin producing unsafe, inaccurate, or policy-violating content.

As enterprises embed LLMs deeper into critical workflows — from code generation to customer communications to financial analysis — the consequences of LLM security failures become business-critical. The companies in this category are building the equivalent of web application firewalls (WAFs) for the AI era. The market is young but moving fast, with significant acquisition activity already reshaping the competitive landscape.

All 11 LLM Application Security Companies

Aim Security
Enterprise AI security platform for securing GenAI adoption including copilots, agents, and homegrown apps. Acquired by Cato Networks Sep 2025.
📍 Tel Aviv, Israel Est. 2023
Aurascape
AI security platform providing visibility and control over enterprise AI application usage with DLP and risk analysis.
📍 Santa Clara, CA Est. 2024
Check Point Software
Global cybersecurity leader. Acquired Lakera (~$300M, 2025) to build end-to-end AI Security Platform for enterprises.
📍 Tel Aviv, Israel Est. 1993
F5 (AI Security)
Application delivery and security company. Acquired CalypsoAI ($180M, Oct 2025) for AI guardrails and red teaming.
📍 Seattle, WA Est. 1996
Guardrails AI
Open-source framework for adding validation and safety guardrails to LLM applications with structured output enforcement.
📍 San Francisco, CA Est. 2023
Harmonic Security
AI security platform providing real-time data protection and visibility for enterprise GenAI usage and shadow AI.
📍 Austin, TX Est. 2023
Lasso Security
GenAI security platform for discovering, monitoring, and securing enterprise LLM usage with runtime guardrails.
📍 Tel Aviv, Israel Est. 2023
Pangea Security
Security-as-a-service platform providing AI Guard, vault, and compliance APIs for developers building AI applications. Acquired by CrowdStrike (Sep 2025).
📍 Palo Alto, CA Est. 2021
Pillar Security
AI security platform providing runtime protection for LLM applications against prompt injection, data leaks, and misuse.
📍 Tel Aviv, Israel Est. 2023
Prompt Security
All-in-one enterprise platform securing employee AI usage and homegrown AI apps. Pending acquisition by SentinelOne (~$250M).
📍 New York, NY Est. 2023
WitnessAI
AI security platform with red teaming (Witness Attack) and AI firewall (Witness Protect) for enterprise LLMs.
📍 San Jose, CA Est. 2023
Related Categories

Explore Adjacent Markets

Explore the Full Database

206 companies across 10 categories — search, filter, and analyze the AI security landscape.

Browse All Companies →