WitnessAI

What WitnessAI does

WitnessAI is a privately held, venture-backed AI security and governance company founded in 2023 and headquartered in Mountain View, California. The company sells an AI-native, network-layer platform that observes, protects, and governs every AI interaction across employees, applications, models, and AI agents — without requiring browser extensions or endpoint clients. The platform is organized around four core modules (Observe for shadow AI discovery across 4,000+ AI applications; Protect, a bidirectional AI firewall that blocks prompt injection, jailbreaks, and model manipulation; Control for intent-based policy enforcement, intelligent LLM routing, and audit trails; and Attack, an automated red-teaming service), extended by the Agentic Control capability for governing AI agents and Model Context Protocol (MCP) servers and packaged as four solution bundles (for Applications, Employees, Developers, and Compliance).

The technology is differentiated by a proprietary intent-based ML classification engine that interprets conversational context rather than relying on keyword/regex pattern matching, plus an MCP Catalog scoring tools against OWASP and CVE risk classes and organization-wide approved-tool enforcement at the network layer. The platform secures text, code, structured data, and multimodal AI interactions across more than 100 supported model types and claims a 99.3% true-positive rate on employee AI guardrails.

WitnessAI sells exclusively to large enterprises (Fortune 500 and global equivalents) via a direct field-sales motion (primary CTA: "Book a Demo"), complemented by strategic channel/reseller partnerships (NTT DATA Japan, TENEX.AI managed services), OEM integrations (F5 BIG-IP, Microsoft Copilot), and an emerging MSSP/VAR channel program. Pricing is custom quote-based with annual contracts; the company has not disclosed public tiers or pricing. Revenue mechanics are enterprise SaaS subscriptions with reported growth of more than 500% in ARR over the trailing 12 months, alongside 5x headcount growth. The company has raised $85.5M across two equity rounds (Series A $27.5M May 2024 led by Ballistic Ventures and GV; Series B $58M January 2026 led by Sound Ventures) at a reported post-money valuation above $450M. Named enterprise customers span airline, financial services, hospitality, retail, wealth management, and telecommunications, with deployments including a Top 5 Global Telecom covering 150,000+ employees in under 10 days and a Fortune 500 retailer securing 3 million weekly AI chatbot conversations.

Name

WitnessAI

Legal name

WitnessAI, Inc.

Website

https://witness.ai

Company type

Private

Founded year

2023

Operating status

Operating

Headcount range

101–250 employees

Short description

WitnessAI sells an AI-native, network-layer security and governance platform that observes, protects, and controls every AI interaction across employees, applications, models, and AI agents for Fortune 500 and global enterprises. The company delivers intent-based ML classification, a bidirectional AI firewall, and agentic/MCP governance via direct enterprise sales and channel partnerships.

Ownership category

akta.pro rank

Product category

AI Security and Governance Software

NAICS

Security Systems Services (except Locksmiths) (561621)

SIC

Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

Responsible AI, Security & Privacy Platforms (Safety, Guardrails, PII) (HDAEANAG)

Keywords

• AI security platform
• AI governance software
• LLM runtime protection
• AI firewall
• Shadow AI discovery

Headquarters

HQ city

Mountain View

HQ country

United States

HQ region

North America

Offices2 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Marketing or Sales, Operations, Infrastructure

Revenue model

1. Enterprise SaaS Subscription (Annual Recurring Revenue): Platform sold as enterprise subscription with reported 500%+ ARR growth over the past 12 months. Deal sizes align with Fortune 500 buyers, with case studies including 150,000+ employee rollouts in under 10 days and deployments serving millions of weekly AI chatbot conversations.
2. Channel/Reseller Revenue (via NTT DATA Japan): NTT DATA Japan acts as strategic reseller of the WitnessAI security platform bundled into its "Responsible and Secure AI" end-to-end AI governance service for Japanese enterprise customers.
3. Managed Security Services Revenue (via TENEX.AI): TENEX.AI partnership delivers managed security and compliance services for enterprise AI, generating revenue through managed service delivery on top of the WitnessAI platform.

Pricing tiers

Go-to-market motion5 records

Distribution channels7 records

Marketing channels8 records

Core offering

WitnessAI sells a unified, AI-native security and governance platform that operates at the network layer to observe, protect, and control every AI interaction across employees, models, applications, and AI agents—without requiring browser extensions or endpoint clients. The platform is composed of the Observe (visibility/shadow AI catalog), Protect (bidirectional AI firewall), Control (intent-based policy and routing), and Attack (automated red-teaming) modules, packaged into solution bundles for Applications, Employees, Developers, and Compliance, and sold as an enterprise SaaS subscription.

Product overview

WitnessAI delivers a single unified AI security and governance platform (the WitnessAI Platform) organized around four core modules — Observe (visibility), Protect (AI firewall), Control (governance/routing), and Attack (automated red teaming) — operating at the network layer to govern both human and AI agent workforces across 4,000+ AI applications. These modules are extended by the Agentic Control capability (governing AI agents, MCP servers, and tools) and packaged as four solution bundles (WitnessAI for Applications, for Employees, for Developers, for Compliance) that target specific enterprise personas. Underneath the modules are named sub-product guardrails (Model Protection, Model Identity Protection, Harmful Response Prevention, Organizational Behavior) and a deployment extension (Witness/Anywhere) for distributed teams.

Differentiator

Problem solved

Functional benefit

Brands

• WitnessAI Agentic Control: Control plane for discovering, governing, and securing AI agents, MCP servers, and tools.
• Witness Protect
• Witness Attack
• Witness Observe
• Witness Control

Products and services

• WitnessAI Unified AI Security and Governance Platform Network-level, single-tenant AI security and governance platform providing visibility, protection, and control over every AI interaction across employees, models, applications, and agents. Secures more than 4,000 AI applications with a claimed 99.3% true positive rate on employee AI guardrails.
• WitnessAI Observe Visibility module that scans the network for AI applications, agents, and MCP servers; catalogs them, visualizes real-time conversations, and classifies intent to surface shadow AI usage and inform policy creation.
• WitnessAI Protect (Enterprise AI Firewall) Bidirectional runtime AI firewall that blocks prompt injections, jailbreaks, and model manipulation before they reach models/agents and filters harmful, off-brand, or sensitive responses before users or downstream agents see them; also tokenizes PII, credentials, and secrets.
• WitnessAI Control Governance module that applies allow/warn/block/route actions using intent-based policies, routes sensitive prompts to approved internal models, enforces approved MCP server/tool lists, and generates granular audit trails for compliance.
• WitnessAI Attack (Automated AI Red Teaming) Automated AI red-teaming module that simulates sophisticated attacks (multimodal attack vectors, multi-step jailbreaks, reinforcement-learning techniques) against models pre-deployment to surface vulnerabilities and provide actionable hardening insights.
• WitnessAI Agentic Control Agent governance extension delivering a single control plane for discovering AI agents, scoring MCP servers/tools against OWASP and CVE risk via an MCP Catalog, enforcing organization-wide approved-tool allow lists, and applying runtime agentic enforcement across IDEs, chat apps, and custom agents.
• WitnessAI Agentic Security Agent-focused security capability that monitors active AI agents, tracks their MCP server and tool access, and blocks malicious prompts before they reach agents, addressing unchecked-permission risk in autonomous AI deployments.
• WitnessAI for Applications Solution bundle providing runtime security for AI applications, models, and agents, including bidirectional defense against prompt injection/jailbreaks, intent-based policy enforcement, and pre-deployment adversarial testing (Attack).
• WitnessAI for Employees Solution bundle providing LLM guardrails for employees: visibility into AI usage, single-tenant data protection, role/intent-based policies, and enforcement of approved MCP server and tool lists for employee-driven agentic activity.
• WitnessAI for Developers Solution bundle securing AI coding tools and coding agents; catalogs developer AI tools, protects proprietary source code from exfiltration, masks PII/credentials in real time, and routes requests to secure internal models without SDK or workflow changes.
• WitnessAI for Compliance Solution bundle for AI compliance and audit readiness: single-tenant isolation with customer-controlled encryption, automated sensitive data detection/masking, granular audit logs, and policy enforcement tailored for regulated industries such as PCI DSS 4.0.1.

Quantifiable outcome

• 99.3% true positive rate on employee AI guardrails
• +9 more outcomes

Named customers7 records

Ideal customer profiles2 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration3 records

AI capability4 records

Feature10 records

Scale indicators13 records

Recent moves7 records

Expansion highlights7 records

Direct peers

• Lakera: Direct peer in LLM and AI application security, offering prompt injection defense, red teaming and guardrails for enterprise GenAI deployments. Closely comparable product scope and enterprise target buyer to WitnessAI's Protect and Attack modules.
• Prompt Security: Direct competitor providing AI security across employees, applications and agents — including visibility, prompt protection and data loss prevention for GenAI use. Overlaps with WitnessAI's Observe, Protect and Control modules in both technology and target customer.
• Calypso AI: Direct peer offering AI security posture management, adversarial testing and runtime guardrails for enterprise LLMs and agents. Comparable product portfolio and regulated-industry customer base (financial services, government) to WitnessAI.
• Protect AI: Direct competitor in AI/ML security covering model scanning, supply-chain risk and LLM guardrails. Overlaps with WitnessAI's red teaming, model protection and AI governance capabilities.
• HiddenLayer: Direct peer providing AI security posture management, adversarial attack defense and threat detection for AI models and applications. Comparable scope to WitnessAI's Attack and Protect modules with similar enterprise buyer focus.

Broad incumbents

• Zscaler: Broad incumbent in SSE/SASE/ZTNA that has launched AI-specific security capabilities within its Zero Trust Exchange. Competes for the same CISO budget as WitnessAI but as part of a much larger platform, putting standalone AI security vendors under pricing pressure.
• Palo Alto Networks (Prisma AIRS): Broad incumbent offering Prisma AIRS for AI runtime security and AI access governance as part of its broader security platform. Overlaps with WitnessAI on AI posture management and guardrails but competes via platform consolidation rather than point solution.
• Netskope: Broad incumbent in SSE and CASB that has extended into AI usage controls for SaaS and GenAI applications. Comparable enterprise buyer and overlapping functionality with WitnessAI's Observe and Control modules.
• Cloudflare: Broad incumbent adding AI security and governance capabilities (e.g., AI Gateway, prompt injection defense) within its global network platform. Competes with WitnessAI's network-level architecture by offering AI controls on top of a much larger CDN/security footprint.
• Cisco (Hypershield / AI Defense): Broad incumbent that has launched AI security offerings (AI Defense, Hypershield) targeting enterprise AI policy enforcement and runtime protection. Comparable enterprise buyer and a potential bundling threat to standalone platforms like WitnessAI.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat7 records

Key risks6 records

Key highlights7 records

Customer concentration

• Facebookfacebook.com/witnessai
• LinkedInlinkedin.com/company/witnessai
• Instagraminstagram.com/witnessai

Compliance1 record

Revenue estimate

Valuation estimate

Number of profiles

Profiles10 records

Funding overview

Funding rounds3 records

Investors14 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about WitnessAI