AI Security Startups Map — Agentic AI Vendors by Category

Observability & Governance

258
Continuous discovery, inventory, and behavioral profiling of every AI agent across the enterprise. Lifecycle controls, audit trails, and policy enforcement to keep the agent fleet visible and accountable.

• Prompt Security

• CyBox Security

• Quint

• Glow

• Aegis Preflight

• Principled Intelligence

• VORTIQ-X Consilium

• Upwind

• EVE Security

• Pluto Security

• Obsidian Security

• Kanopy Security

• Onyx Security

• Lumia Security

• Helmet Security

• Runlayer

• Bay Security

• Quilr AI

• JetStream Security

• Straiker

• Token Security

• General Analysis

• Manifold Security

• Relyance AI

• Knostic

• Zenity

• Noma Security

• Aurascape

• WitnessAI

• Clutch Security

• Pillar Security

• Cranium AI

• Fiddler AI

• Willow

• Mindgard

• MintMCP

• Archestra

• Trent AI

• Singulr AI

• Tumeryk

• Canyon Road

• Geordie AI

• Realm Labs

• Agentic Fabriq

• Certiv

• Golf

• Entro Security

• Oryo

• Cymphony

• Opsin Security

• Bonfy

• Portal26

• Grip Security

• SurePath AI

• Ray Security

• Vorlon

• SlashID

• Permiso

• ThirdLaw

• Galileo

• Openlayer

• Obot AI

• Valence Security

• Reco

• Opti

• Twine Security

• Backslash

• PointGuard AI

• Autonomous

• Island

• Keyboard

• Polaxis

• Enkrypt AI

• Requesty

• Future AGI

• SlashLLM

• CodeIntegrity

• Pallma

• Dash Security

• Raven

• Eos Cyber

• Nestria AI

• Sonoma Security

• Lunar.dev

• HikmaAI

• NuDay AI

• raxIT Labs

• Assury

• Stacklok

• Oblivious

• Lumeus

• IndyKite

• NeuralTrust

• Parakeet Security

• Virtue AI

• Aira Security

• DeTaSECURE

• Aiceberg

• Tetrate

• Pynt

• Permit.io

• Edera

• Wallarm

• CloudMatos

• Adversa AI

• Oasis Security

• Trust3 AI

• Nexos.ai

• Trustwise

• FireTail

• Riscosity

• Calvin Risk

• FairNow

• Zendata

• Confident Security

• Aona AI

• FireRaven

• Galene.AI

• Cygeniq AI

• KonaSense

• AI Score

• AI Security Gateway

• Secuvy

• PrivaSapien

• Acuvity

• Credo AI

• Daxa.ai

• Oso

• Modulos

• Safe Intelligence

• Aurva

• Teleskope

• Advai

• Crash Override

• LangGuard

• SolidCore

• Eroun & Company

• Aim Security

• Apex Security

• Pangea

• Portkey

• Koi

• Astrix

• Genie Security

• Ryft

• Robust Intelligence

• Trail Security

• SPLX

• Symmetry Systems

• CalypsoAI

• Protect AI

• LayerX Security

• Deepchecks

• Promptfoo

• Cyata

• Airrived

• Unbound

• TestSavant

• Verax

• Cyera

• SUPERWISE

• First Recon AI

• Tego AI

• Islo

• DapplePot

• Reva

• Rockfort AI

• Tenet Security

• Origin

• Red Access

• Anomity

• MagicMirror

• Polygraf AI

• Citadel AI

• BlueRock

• AGAT Software

• Airia

• Arthur

• Ciphero

• Credal

• AllTrue.ai

• Guardion AI

• Identity Machines

• Intentyx

• Kipling Secure

• Liminal Security

• Modelmetry

• Nroc Security

• Swift Security

• Tibo

• Unseen Security

• Verosek

• Intellect Machines

• Waxell

• Optimus Labs

• EqtyLab

• Xyra Security

• Fortifai

• Natoma

• Barndoor AI

• Nokod Security

• Holistic AI

• ModelOp

• Monitaur

• Saidot

• Lumenova AI

• Fairly AI

• C1

• Patronus AI

• Sentra

• Arize AI

• AstraSync AI

• ActiveFence

• Styrk

• Maro

• ChatSee

• Jazz Security

• NewCore

• Keep Aware

• Ent

• LangProtect

• WideField

• Repacket

• Safety

• AccuKnox

• Secure Agentics

• A Security

• AuthMind

• Silverfort

• 8Layers

• Oleria

• Redblock

• Offroad

• AKA Identity

• Fabrix

• aizome

• Oak

• Pallas Security

• Senthex

• Neo

• Bloom

• Push Security

• Burgus

• KIPIO

• Arrakis Security

• MCP & LLM Gateways

• 89
  Protocol-layer control plane sitting between agents and the MCP servers, tools, and models they call. Per-request auth, scope enforcement, tool-call inspection, and supply-chain governance for the agent ↔ tool boundary.

• Prompt Security

• CyBox Security

• AuthPlane

• Principled Intelligence

• Helmet Security

• Operant AI

• Runlayer

• Straiker

• Lasso Security

• Descope

• agen.co by Frontegg

• Willow

• MintMCP

• Archestra

• Arcade.dev

• Agentic Fabriq

• Golf

• Aembit

• Oryo

• Harmon

• Grip Security

• Vorlon

• SlashID

• Permiso

• Cequence Security

• Pomerium

• PVML

• Openlayer

• Obot AI

• PipeLab

• PointGuard AI

• Autonomous

• Island

• Keyboard

• Polaxis

• CodeIntegrity

• Eos Cyber

• Lunar.dev

• NuDay AI

• Assury

• Oblivious

• NeuralTrust

• Permit.io

• Oasis Security

• Oso

• authID

• Pangea

• Portkey

• Astrix

• Symmetry Systems

• Cyata

• Islo

• Reva

• Alter

• ArmorIQ

• EnforceAuth

• Identity Machines

• Gopher Security

• Skyrelis

• DeepTrail

• Natoma

• Barndoor AI

• Beltic

• C1

• AstraSync AI

• Hush Security

• GraphnAI

• NewCore

• Apono (acquired by 1Password)

• Mitiga

• WideField (acquired by Cisco)

• AccuKnox

• AuthMind

• Silverfort

• 8Layers

• Oleria

• Redblock

• Offroad

• AKA Identity

• Fabrix (acquired by Silverfort)

• aizome

The landscape by the numbers

Live figures computed from the dataset behind this map. Funding reflects disclosed rounds compiled from public announcements.

• 402 AI security vendors tracked
• 14 functional categories
• 27 countries represented
• $11.3B in disclosed funding

Frequently asked questions

How many AI security startups are there?

The AI Security Startups Map tracks 402 companies building security for agentic AI, organized across 14 functional categories and spanning 27 countries. The directory includes acquired companies whose products still ship and excludes stealth-mode startups.

What are the main categories of AI security vendors?

Vendors are grouped into 14 categories. The largest is Observability & Governance with 258 vendors, followed by Runtime & Guardrails (203) and Agentic Identity (93). Other categories include AI red teaming, MCP & LLM gateways, agentic identity, model security, and AI-SPM.

Where are AI security companies based?

The 402 vendors are headquartered across 27 countries. The top three by company count are USA (214), Israel (90), UK (20). The United States and Israel together account for the majority of the landscape.

How much funding have AI security startups raised?

The landscape has raised roughly $11.3B in total. Funding figures reflect disclosed rounds compiled from public announcements.

When were most AI security startups founded?

Founding activity is heavily concentrated in recent years: 232 of the 402 vendors (58%) were founded between 2023 and 2025, with 2025 the single busiest year at 87 new companies.

Which companies are acquiring AI security startups?

35 vendors on the map have been acquired. The most active consolidators are Check Point (3), PAN (3), Cisco (3).

Are all of these companies new AI-native startups?

No. 118 of 402 vendors (29%) are pre-existing security companies that repositioned into AI/agentic security as the market grew.

Published June 5, 2025 · Last updated August 31, 2026.