Developer docs
API playgroundTry for free, no card

Search company profiles

Adversa AI

Full company profile

uuid0003cfz

Namestring
Adversa AI
Legal namestring
Adversa AI
Websiteurl
adversa.ai
Company typeenum
Private
Founded yearint
2019
Descriptiontext

Adversa AI is an Israeli AI cybersecurity company founded in 2019 and headquartered in Tel Aviv that provides continuous red teaming and remediation services for AI systems, with a primary focus on autonomous AI agents, large language models, and generative AI applications. The company's core platform, the Agentic AI Security Platform, runs autonomous security assessment campaigns on every model update, prompt change, and new tool connection, identifying vulnerabilities such as prompt injection, goal hijacking, tool misuse, privilege abuse, and memory poisoning. The company supplements this with SecureClaw (an open-source OWASP-aligned security plugin for OpenClaw AI agents), AIRQ (an independent risk scoring framework that rates 100+ AI agents on attack surface, blast radius, and defenses), and targeted red-teaming service lines for MCP, GenAI, and LLM deployments.

The firm's core technology maps assessments to OWASP, NIST, CSA, and MITRE ATLAS frameworks, and is backed by what the company describes as the world's first AI security patent. Its research team has produced high-profile disclosures including the Universal LLM Jailbreak paper, the SymJack symlink-hijack RCE vulnerability affecting six AI coding agents (Claude Code, Cursor, Antigravity, Copilot, Grok Build), and the TrustFall MCP auto-approval flaw, which are evidence of meaningful adversarial AI expertise rather than surface-level tooling. Adversa AI co-leads the CoSAI Agentic AI Security workstream and contributes to standards bodies including OWASP, NIST, and CSA, giving it influence over emerging AI security compliance requirements.

The company operates a B2B enterprise GTM with two revenue streams: subscription-based access to the continuous red-teaming platform and managed professional services for hands-on security assessment. Pricing is quote-based and not publicly disclosed. Target customers are enterprise organizations across financial services, insurance, automotive, internet/media/marketplaces, and smart city verticals that are deploying AI agents for high-stakes use cases such as trading, claims processing, and autonomous driving. The GTM motion is anchored in thought leadership through research publication, standards-body contribution, conference speaking (including RSA Conference 2026), and earned media presence in outlets including Fortune, Forbes, MIT Technology Review, WIRED, TechCrunch, and the Wall Street Journal.

Short descriptiontext

Adversa AI is an Israeli AI cybersecurity company providing continuous red teaming and remediation for autonomous AI agents, LLMs, and generative AI applications, serving enterprise customers across financial services, insurance, automotive, internet/media, and smart city verticals.

Operating statusenum
Operating
Ownership categoryenum
Headcount rangeband
11–50
akta.pro rankint
HeadquartersTel Aviv, Israel
HQ citystring
Tel Aviv
HQ countrystring
Israel
HQ regionstring
Middle East
Markets served

Serves global market

Offices1 record

Each record includes

City, Country, Type, Description, Source

Keyword5 values
AI red teaming, AI security platform, LLM vulnerability testing, Agentic AI security, AI threat modeling
Industry1 code
PRIMARYHDAAAKAC
Model Security Testing & Red Teaming (adversarial ML, jailbreaks)
NAICS code1 code
  • Custom Computer Programming Services541511
SIC code2 codes
  • Services-Prepackaged Software7372
  • Services-Computer Programming, Data Processing, Etc.7370
Product category
AI Security Software
Social media profiles2 records
GTM motion1 record

Each record includes

Type, Description, Source

Revenue model3 records
1AI Red Teaming Platform
TypeSubscription Recurring
Description

Continuous red teaming platform subscription for enterprises to test and secure AI agents, LLMs, and GenAI applications. Includes autonomous security assessment campaigns.

adversa.ai
2Managed Services
TypeProfessional Services
Description

Managed service offering providing professional AI security assessment and advisory services for organizations needing hands-on support.

adversa.ai
3Enterprise Security Assessment
TypeSubscription Recurring
Description

Custom threat modeling and security assessments for specific AI implementations including financial, insurance, automotive, and other industry verticals.

prnewswire.com
Marketing channels6 records

Each record includes

Title, Type, Stage, Description, Source

Distribution channels2 records

Each record includes

Title, Type, Scope, Target buyer, Description, Source

Cost components4 values
Personnel, Technology or R&D, Marketing or Sales, Operations
GTM typeB2B
B2B
Offering typeSoftware
Software
Brand1 of 3 records shown
1SecureClaw
Description

Open-source security plugin and behavioral security skill designed specifically to secure OpenClaw AI agents, aligned with OWASP standards and covering threat categories such as MITRE ATLAS.

prnewswire.com
+2 more records
Core offering1 text field

Adversa AI delivers continuous AI red teaming and remediation for custom AI agents, LLMs, and GenAI applications via its Agentic AI Security Platform. The platform autonomously runs security assessments on every model update, prompt change, and new tool connection, identifying vulnerabilities such as prompt injection, goal hijacking, tool misuse, privilege abuse, and memory poisoning. It also offers specialized modules (SecureClaw, AIRQ), targeted red teaming for MCP/GenAI/LLM, and tiered services (Secure AI Awareness, Secure AI Assessment, Secure AI Assurance).

Differentiator
Functional benefit
Problem solved
Product overview1 text field

Adversa AI offers a unified security platform for AI systems consisting of the core Agentic AI Security Platform with continuous red teaming and remediation capabilities, complemented by specialized modules including SecureClaw (open-source OWASP-aligned security plugin for OpenClaw agents), AIRQ (AI Risk Quadrant Report for ranking 100+ AI agents), and targeted red teaming services for MCP, GenAI, and LLM. The platform is structured around three service tiers: Secure AI Awareness (threat intelligence and training), Secure AI Assessment (vulnerability identification), and Secure AI Assurance (remediation and monitoring). The company pioneered AI security with patent-pending technology and contributes to OWASP, NIST, and CSA standards.

Product and service9 records
1Agentic AI Security Platform
CategoryCore Platform
Description

Continuous AI red teaming and remediation platform for custom AI agents. Runs autonomous security assessment campaigns on every model update, prompt change, and new tool connection, identifying vulnerabilities such as prompt injection, goal hijacking, tool misuse, privilege abuse, and memory poisoning, then auto-generates patches and actionable reports for prioritized remediation. For enterprises deploying autonomous AI agents and LLMs.

2AI Red Teaming for MCP
CategorySpecialized Red Teaming Module
Description

Security assessment and red teaming services specifically designed for Model Context Protocol (MCP) implementations, addressing unique vulnerabilities in AI tool integration frameworks. For enterprises deploying MCP-based AI agents and tool integrations.

3AI Red Teaming for GenAI
CategorySpecialized Red Teaming Module
Description

Security testing and vulnerability assessment for generative AI applications and systems, covering prompt injection, data leakage, and model manipulation risks. For organizations deploying generative AI applications in production.

4AI Red Teaming for LLM
CategorySpecialized Red Teaming Module
Description

Comprehensive security testing for large language models including jailbreak evaluation, prompt leaking detection, and adversarial attack assessment. For enterprises deploying LLMs in production.

5SecureClaw for OpenClaw
CategoryOpen-Source Security Module
Description

Open-source OWASP-aligned security plugin and behavioral security skill for OpenClaw AI agents. Provides automated auditing, hardening, and behavioral checks addressing the full attack surface per OWASP standards and MITRE ATLAS threat categories including prompt injection, credential exposure, and supply-chain threats. For OpenClaw AI agent developers and security teams.

6AIRQ (AI Risk Quadrant Report)
CategoryRisk Assessment Framework
Description

First independent AI agent security rating and open-source risk scoring framework. Ranks 100+ AI agents across attack surface, blast radius, and defenses. Enables enterprises to compare AI agents by security risk and assess their own AI stack using the open-source methodology. For enterprise security teams evaluating AI agent deployments.

7Secure AI Awareness
CategoryManaged Advisory Service
Description

Service to demonstrate AI risks and shape AI governance strategy, consisting of Policy Checkup, AI Risks Training, and Threat Intelligence for informed decisions. For organizations building their initial AI security awareness and governance programs.

8Secure AI Assessment
CategoryManaged Assessment Service
Description

AI integrity validation and vulnerability identification through Threat Modeling, Vulnerability Audit, and automated AI Red Teaming. For organizations needing a structured AI vulnerability assessment engagement.

9Secure AI Assurance
CategoryManaged Assurance Service
Description

AI risk remediation and lifecycle management for AI integrity, consisting of Security Evaluation, Risk Mitigation, and Attack Detection capabilities. For organizations requiring ongoing AI security assurance after deployment.

Scale indicator3 records

Each record includes

Type, Value, Description, Source

Partnership5 partners
Strategic tierCoreTypeStrategic or Co-development Partner
Description

Adversa AI co-leads the CoSAI Agentic AI Security workstream and actively contributes to OWASP AI security frameworks including the Agentic AI Security Top 10. Company experts help author OWASP standards and reference guides.

Strategic tierCoreTypeStrategic or Co-development Partner
Description

Contributes to NIST AI security standards development as part of the company's thought leadership and industry standards engagement.

3CSA (Cloud Security Alliance)
Strategic tierCoreTypeStrategic or Co-development Partner
Description

Contributes to CSA AI security standards and participated in CSA Pitchapalooza 2025 award competition.

prnewswire.com
4CoSAI (Coalition for Secure AI)
Strategic tierCoreTypeStrategic or Co-development Partner
Description

Co-leads the Agentic AI Security workstream within CoSAI, contributing to the development of AI security frameworks and best practices.

prnewswire.com
Strategic tierCoreTypeTechnology or Integration
Description

SecureClaw, developed by Adversa AI, is an open-source OWASP-aligned security plugin and skill specifically designed for OpenClaw AI agents. Company plans to continue developing security solutions for the OpenClaw ecosystem.

Recent move6 records

Each record includes

Date, Type, Title, Description, Source

Expansion highlight5 records

Each record includes

Type, Description

Peers10 records
TypeDirect peer
Description

Direct competitor offering AI/ML model security, red teaming, and adversarial attack protection for enterprise AI systems. Closely comparable product category, target market, and use cases (LLM and ML model protection).

TypeDirect peer
Description

Competitor providing AI security testing and adversarial ML assessment for enterprise AI systems. Comparable red teaming and vulnerability research positioning.

TypeDirect peer
Description

Competitor in AI/ML model security with focus on detecting trojans, backdoors, and adversarial vulnerabilities in models. Adjacent but overlapping offering in enterprise AI security validation.

TypeDirect peer
Description

Competitor focused on AI security testing and adversarial defense for enterprise LLM deployments. Comparable offering in red teaming and model validation for generative AI.

TypeDirect peer
Description

Competitor in AI/ML security with red teaming, scanning, and threat modeling for LLMs and ML models. Recently acquired by Palo Alto Networks, indicating incumbent consolidation of the same category Adversa AI operates in.

TypeDirect peer
Description

Direct competitor in AI model security and validation for enterprise AI deployments. Acquired by Cisco, illustrating the strategic value of the AI security category to large platform vendors.

TypeDirect peer
Description

Competitor offering LLM security, prompt injection defense, and AI red teaming. Overlapping product portfolio in protecting enterprise generative AI applications.

TypeBroad incumbent
Description

Major platform incumbent offering AI red teaming services, Azure AI Content Safety, and Prompt Shields as part of broader Azure AI portfolio. Represents bundled competition for enterprise buyers.

TypeBroad incumbent
Description

Incumbent that acquired Robust Intelligence to embed AI security into its broader networking and security platform. Represents strategic competitor bundling AI red teaming into enterprise security suites.

TypeBroad incumbent
Description

Incumbent offering enterprise AI governance, risk, and compliance tooling via watsonx.governance. Competes in the AI GRC category with broader portfolio reach into regulated enterprises.

Market position
Strengths4 records

Each record includes

Headline, Details, Source

Weaknesses4 records

Each record includes

Headline, Details, Source

Competitive moat5 records

Each record includes

Type, Details

Key risks6 records

Each record includes

Headline, Details, Source

Key highlights6 records

Each record includes

Headline, Details, Source

Customer concentration

Classification, Details

Named customers2 records

Each record includes

Name, Industry, Type, Use case, Source, UUID

Segment5 records

Each record includes

Title, Type, Primary, Description, Pain point addressed, Use case, Source

Ideal customer profile3 records

Each record includes

Profile, Firmographic size, Sales motion, Sales cycle length, Buying structure, Purchase trigger, Buyer persona, Geography, Industry vertical, Primary use case, Description, Pain points, Evidence proof points, Target buyer

Technology focused
Yes
API detail
Has APIbool
No

Docs URL, Description

AI capability6 records

Each record includes

Type, Description, Source

AI maturity
App detail

Has app

Feature6 records

Each record includes

Title, Differentiator, Description, Source

Core technology
Revenue estimate
Valuation estimate
Number of profiles
Profiles2 records

Each record includes

Name, Designation, Designation category, Overview, Profile commentary, Source

No data
Compliance1 record

Each record includes

Name, Class, Description

Funding overview

Funding stage, Last funding date, Total funding USD

Funding rounds1 record

Each record includes

Round, Amount USD, Date, Pre money valuation, Total investors, Investors, News

Investors

Each record includes

Name, Type, Date of entry, Rounds participated, Website

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

M&A

Each record includes

Name, Acquisition type, Announced date, Completed date, Status, Website, News

Investment

Each record includes

Name, Round, Announced date, Lead investor, Website, News

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

Adversa AI

AI Security Softwareadversa.ai

Adversa AI is an Israeli AI cybersecurity company providing continuous red teaming and remediation for autonomous AI agents, LLMs, and generative AI applications, serving enterprise customers across financial services, insurance, automotive, internet/media, and smart city verticals.

What Adversa AI does

Adversa AI is an Israeli AI cybersecurity company founded in 2019 and headquartered in Tel Aviv that provides continuous red teaming and remediation services for AI systems, with a primary focus on autonomous AI agents, large language models, and generative AI applications. The company's core platform, the Agentic AI Security Platform, runs autonomous security assessment campaigns on every model update, prompt change, and new tool connection, identifying vulnerabilities such as prompt injection, goal hijacking, tool misuse, privilege abuse, and memory poisoning. The company supplements this with SecureClaw (an open-source OWASP-aligned security plugin for OpenClaw AI agents), AIRQ (an independent risk scoring framework that rates 100+ AI agents on attack surface, blast radius, and defenses), and targeted red-teaming service lines for MCP, GenAI, and LLM deployments.

The firm's core technology maps assessments to OWASP, NIST, CSA, and MITRE ATLAS frameworks, and is backed by what the company describes as the world's first AI security patent. Its research team has produced high-profile disclosures including the Universal LLM Jailbreak paper, the SymJack symlink-hijack RCE vulnerability affecting six AI coding agents (Claude Code, Cursor, Antigravity, Copilot, Grok Build), and the TrustFall MCP auto-approval flaw, which are evidence of meaningful adversarial AI expertise rather than surface-level tooling. Adversa AI co-leads the CoSAI Agentic AI Security workstream and contributes to standards bodies including OWASP, NIST, and CSA, giving it influence over emerging AI security compliance requirements.

The company operates a B2B enterprise GTM with two revenue streams: subscription-based access to the continuous red-teaming platform and managed professional services for hands-on security assessment. Pricing is quote-based and not publicly disclosed. Target customers are enterprise organizations across financial services, insurance, automotive, internet/media/marketplaces, and smart city verticals that are deploying AI agents for high-stakes use cases such as trading, claims processing, and autonomous driving. The GTM motion is anchored in thought leadership through research publication, standards-body contribution, conference speaking (including RSA Conference 2026), and earned media presence in outlets including Fortune, Forbes, MIT Technology Review, WIRED, TechCrunch, and the Wall Street Journal.

Adversa AI firmographics

Firmographics
Name
Adversa AI
Legal name
Adversa AI
Website
https://adversa.ai
Company type
Private
Founded year
2019
Operating status
Operating
Headcount range
11–50 employees
Short description
Adversa AI is an Israeli AI cybersecurity company providing continuous red teaming and remediation for autonomous AI agents, LLMs, and generative AI applications, serving enterprise customers across financial services, insurance, automotive, internet/media, and smart city verticals.
Ownership category
akta.pro rank

Adversa AI industry classification

Industry
Product category
AI Security Software
NAICS
Custom Computer Programming Services (541511)
SIC
Services-Prepackaged Software (7372), Services-Computer Programming, Data Processing, Etc. (7370)
akta.pro industry
PRIMARYHDAAAKAC
Model Security Testing & Red Teaming (adversarial ML, jailbreaks)

Keywords

  • AI red teaming
  • AI security platform
  • LLM vulnerability testing
  • Agentic AI security
  • AI threat modeling

Where Adversa AI is headquartered

Location

Headquarters

HQ city
Tel Aviv
HQ country
Israel
HQ region
Middle East

Offices1 record

Markets served

Adversa AI business model

Business model
GTM type
B2B
Offering type
Software
Cost components
Personnel, Technology or R&D, Marketing or Sales, Operations

Revenue model

  1. AI Red Teaming Platform: Continuous red teaming platform subscription for enterprises to test and secure AI agents, LLMs, and GenAI applications. Includes autonomous security assessment campaigns.
  2. Managed Services: Managed service offering providing professional AI security assessment and advisory services for organizations needing hands-on support.
  3. Enterprise Security Assessment: Custom threat modeling and security assessments for specific AI implementations including financial, insurance, automotive, and other industry verticals.

Go-to-market motion1 record

Distribution channels2 records

Marketing channels6 records

Adversa AI product offering

Product offering

Core offering

Adversa AI delivers continuous AI red teaming and remediation for custom AI agents, LLMs, and GenAI applications via its Agentic AI Security Platform. The platform autonomously runs security assessments on every model update, prompt change, and new tool connection, identifying vulnerabilities such as prompt injection, goal hijacking, tool misuse, privilege abuse, and memory poisoning. It also offers specialized modules (SecureClaw, AIRQ), targeted red teaming for MCP/GenAI/LLM, and tiered services (Secure AI Awareness, Secure AI Assessment, Secure AI Assurance).

Product overview

Adversa AI offers a unified security platform for AI systems consisting of the core Agentic AI Security Platform with continuous red teaming and remediation capabilities, complemented by specialized modules including SecureClaw (open-source OWASP-aligned security plugin for OpenClaw agents), AIRQ (AI Risk Quadrant Report for ranking 100+ AI agents), and targeted red teaming services for MCP, GenAI, and LLM. The platform is structured around three service tiers: Secure AI Awareness (threat intelligence and training), Secure AI Assessment (vulnerability identification), and Secure AI Assurance (remediation and monitoring). The company pioneered AI security with patent-pending technology and contributes to OWASP, NIST, and CSA standards.

Differentiator

Problem solved

Functional benefit

Brands

  • SecureClaw: Open-source security plugin and behavioral security skill designed specifically to secure OpenClaw AI agents, aligned with OWASP standards and covering threat categories such as MITRE ATLAS.
  • AIRQ
  • Agentic AI Security Platform

Products and services

  • Agentic AI Security Platform Continuous AI red teaming and remediation platform for custom AI agents. Runs autonomous security assessment campaigns on every model update, prompt change, and new tool connection, identifying vulnerabilities such as prompt injection, goal hijacking, tool misuse, privilege abuse, and memory poisoning, then auto-generates patches and actionable reports for prioritized remediation. For enterprises deploying autonomous AI agents and LLMs.
  • AI Red Teaming for MCP Security assessment and red teaming services specifically designed for Model Context Protocol (MCP) implementations, addressing unique vulnerabilities in AI tool integration frameworks. For enterprises deploying MCP-based AI agents and tool integrations.
  • AI Red Teaming for GenAI Security testing and vulnerability assessment for generative AI applications and systems, covering prompt injection, data leakage, and model manipulation risks. For organizations deploying generative AI applications in production.
  • AI Red Teaming for LLM Comprehensive security testing for large language models including jailbreak evaluation, prompt leaking detection, and adversarial attack assessment. For enterprises deploying LLMs in production.
  • SecureClaw for OpenClaw Open-source OWASP-aligned security plugin and behavioral security skill for OpenClaw AI agents. Provides automated auditing, hardening, and behavioral checks addressing the full attack surface per OWASP standards and MITRE ATLAS threat categories including prompt injection, credential exposure, and supply-chain threats. For OpenClaw AI agent developers and security teams.
  • AIRQ (AI Risk Quadrant Report) First independent AI agent security rating and open-source risk scoring framework. Ranks 100+ AI agents across attack surface, blast radius, and defenses. Enables enterprises to compare AI agents by security risk and assess their own AI stack using the open-source methodology. For enterprise security teams evaluating AI agent deployments.
  • Secure AI Awareness Service to demonstrate AI risks and shape AI governance strategy, consisting of Policy Checkup, AI Risks Training, and Threat Intelligence for informed decisions. For organizations building their initial AI security awareness and governance programs.
  • Secure AI Assessment AI integrity validation and vulnerability identification through Threat Modeling, Vulnerability Audit, and automated AI Red Teaming. For organizations needing a structured AI vulnerability assessment engagement.
  • Secure AI Assurance AI risk remediation and lifecycle management for AI integrity, consisting of Security Evaluation, Risk Mitigation, and Attack Detection capabilities. For organizations requiring ongoing AI security assurance after deployment.

Companies that use Adversa AI

Customer profile

Named customers2 records

Segments5 records

Ideal customer profiles3 records

Adversa AI technology and API

Technology

Technology focussed Yes

API detail

Has API
No
API docs
API detail

Core technology

AI maturity

App detail

AI capability6 records

Feature6 records

Adversa AI partnerships and signals

Strategic signal

Partnerships

Five partnerships are on record, tiered core.

  • OWASPcoreStrategic or Co-development PartnerAdversa AI co-leads the CoSAI Agentic AI Security workstream and actively contributes to OWASP AI security frameworks including the Agentic AI Security Top 10. Company experts help author OWASP standards and reference guides.
  • NISTcoreStrategic or Co-development PartnerContributes to NIST AI security standards development as part of the company's thought leadership and industry standards engagement.
  • CSA (Cloud Security Alliance)coreStrategic or Co-development PartnerContributes to CSA AI security standards and participated in CSA Pitchapalooza 2025 award competition.
  • CoSAI (Coalition for Secure AI)coreStrategic or Co-development PartnerCo-leads the Agentic AI Security workstream within CoSAI, contributing to the development of AI security frameworks and best practices.
  • OpenClawcoreTechnology or IntegrationSecureClaw, developed by Adversa AI, is an open-source OWASP-aligned security plugin and skill specifically designed for OpenClaw AI agents. Company plans to continue developing security solutions for the OpenClaw ecosystem.

Scale indicators3 records

Recent moves6 records

Expansion highlights5 records

Adversa AI competitors and assessment

Company assessment

Direct peers

  • HiddenLayer: Direct competitor offering AI/ML model security, red teaming, and adversarial attack protection for enterprise AI systems. Closely comparable product category, target market, and use cases (LLM and ML model protection).
  • Mindgard: Competitor providing AI security testing and adversarial ML assessment for enterprise AI systems. Comparable red teaming and vulnerability research positioning.
  • TrojAI: Competitor in AI/ML model security with focus on detecting trojans, backdoors, and adversarial vulnerabilities in models. Adjacent but overlapping offering in enterprise AI security validation.
  • Calypso AI: Competitor focused on AI security testing and adversarial defense for enterprise LLM deployments. Comparable offering in red teaming and model validation for generative AI.
  • Protect AI: Competitor in AI/ML security with red teaming, scanning, and threat modeling for LLMs and ML models. Recently acquired by Palo Alto Networks, indicating incumbent consolidation of the same category Adversa AI operates in.
  • Robust Intelligence: Direct competitor in AI model security and validation for enterprise AI deployments. Acquired by Cisco, illustrating the strategic value of the AI security category to large platform vendors.
  • Lakera: Competitor offering LLM security, prompt injection defense, and AI red teaming. Overlapping product portfolio in protecting enterprise generative AI applications.

Broad incumbents

  • Microsoft (Azure AI Red Team / AI Content Safety): Major platform incumbent offering AI red teaming services, Azure AI Content Safety, and Prompt Shields as part of broader Azure AI portfolio. Represents bundled competition for enterprise buyers.
  • Cisco (post Robust Intelligence acquisition): Incumbent that acquired Robust Intelligence to embed AI security into its broader networking and security platform. Represents strategic competitor bundling AI red teaming into enterprise security suites.
  • IBM (watsonx.governance): Incumbent offering enterprise AI governance, risk, and compliance tooling via watsonx.governance. Competes in the AI GRC category with broader portfolio reach into regulated enterprises.

Market position

Strengths4 records

Weaknesses4 records

Competitive moat5 records

Key risks6 records

Key highlights6 records

Customer concentration

Adversa AI social profiles

Digital presence

Adversa AI compliance and trust

Trust signal

Compliance1 record

Adversa AI financial estimates

Financial estimate

Revenue estimate

Valuation estimate

Adversa AI leadership team

Management profile

Number of profiles

Profiles2 records

Adversa AI funding detail

Funding detail

Funding overview

Funding rounds1 record

Investors

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

Adversa AI M&A and investment

M&A and investment

M&A

Investments

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

Frequently asked questions about Adversa AI

What does Adversa AI do?

Adversa AI delivers continuous AI red teaming and remediation for custom AI agents, LLMs, and GenAI applications via its Agentic AI Security Platform. The platform autonomously runs security assessments on every model update, prompt change, and new tool connection, identifying vulnerabilities such as prompt injection, goal hijacking, tool misuse, privilege abuse, and memory poisoning. It also offers specialized modules (SecureClaw, AIRQ), targeted red teaming for MCP/GenAI/LLM, and tiered services (Secure AI Awareness, Secure AI Assessment, Secure AI Assurance).

Is Adversa AI a public or private company?

Adversa AI is a private company. It is classified as founder individual operated bootstrapped and is currently operating.

When was Adversa AI founded?

Adversa AI was founded in 2019. It employs 11 to 50 people.

Where is Adversa AI based?

Adversa AI is headquartered in Tel Aviv, Israel, in the Middle East region.

How does Adversa AI make money?

Three revenue lines are on record. AI Red Teaming Platform is the primary driver. The others are managed Services and enterprise Security Assessment.

Who are Adversa AI's main competitors?

Direct peers on record are HiddenLayer, Mindgard, TrojAI, Calypso AI, Protect AI, Robust Intelligence and Lakera. Broad incumbents are Microsoft (Azure AI Red Team / AI Content Safety), Cisco (post Robust Intelligence acquisition) and IBM (watsonx.governance).

Does Adversa AI have an API?

No public API is recorded for Adversa AI.

What industry is Adversa AI in?

Adversa AI's product category is AI Security Software. Its primary akta.pro industry code is HDAAAKAC, Model Security Testing & Red Teaming (adversarial ML, jailbreaks). Its NAICS code is 541511 and its SIC code is 7372.

Unlock the full company data

50 free credits on sign-up, no credit card required.

Contact sales
Live signals
PR NewswireAdversa AI Launches Coding Agent Security Platform to Block Cyberattacks on AI Agents in Real TimeAdversa AI launched its Coding Agent Security Platform for developer endpoints, blocking attack chains in under 0.2 milliseconds. The platform uses in-memory engines and over 300 default policies, detecting attacks at a 0.1% false positive rate. It was tested at a financial institution, stopping an attack that attempted to exfiltrate API secrets.InfoWorldEncrypted instructions trick Copilot CLI into spilling developer secretsResearchers at Adversa AI demonstrated that GitHub Copilot CLI can read sensitive local files and send their contents to an attacker-controlled endpoint via encrypted instructions. The attack, called Cryptographic Context Injection, requires Copilot CLI in autopilot mode and a model willing to execute the decrypted instructions. GitHub validated the finding but declined to treat it as a vulnerability.CSO OnlineEncrypted instructions trick Copilot CLI into spilling developer secretsResearchers at Adversa AI demonstrated that GitHub Copilot CLI can read sensitive local files and send their contents to an attacker-controlled endpoint via encrypted instructions. The attack, called Cryptographic Context Injection, requires Copilot CLI in autopilot mode and a model willing to execute decrypted instructions. GitHub validated the finding but declined to treat it as a vulnerability.Cyber Security NewsGrok Zero-Click Attack Steals Chat Data Using Encrypted Prompt InjectionAdversa AI disclosed a zero-click vulnerability in xAI's Grok web chat that uses encrypted prompt injection to steal user identity, location, and conversation history. The attack exploits the model's ability to decrypt and execute code within its sandbox, bypassing static input filters without user interaction or warning.SecurityWeekEncrypted Prompts Bypass AI Safety Guardrails in Grok and GeminiResearchers at Adversa AI disclosed a new attack technique called Cryptographic Context Injection that bypasses safety guardrails in xAI's Grok and Google's Gemini by encrypting malicious prompts. The method allows attackers to execute harmful instructions within the models' code execution sandboxes, enabling data exfiltration and the generation of restricted content without triggering standard filters. While xAI has not responded to disclosure attempts, Google's system showed reduced vulnerability by August, prompting the researchers to publish their findings publicly.CoinMarketCapEncrypted web page payload turns Grok into a chat history leak: Guest Post by Cryptopolitan_NewsxAI has failed to patch a vulnerability in its Grok chatbot that allows hackers to leak users' private chat histories via encrypted web page payloads. The exploit, disclosed by Adversa AI in June 2026, bypasses safety filters by injecting commands that Grok decrypts and executes, sending data to attacker-controlled servers.The Hacker NewsNew Cryptographic Context Injection Attack Could Let Web Pages Steal Grok Chat DataAdversa AI disclosed a cryptographic context injection attack that allows xAI's Grok chatbot to exfiltrate user data, including names, locations, and conversation history, to attacker-controlled servers via encrypted web page payloads. The vulnerability affects the Grok 4.5 Fast model and remains unpatched as xAI has not provided a mitigation timeline since being notified in June 2026.The New StackResearchers hid an attack inside AES encryption. The AI model cracked it open willingly.Researchers at Adversa demonstrated Cryptographic Context Injection against xAI's Grok, using AES-256-GCM encrypted payloads to bypass security filters. The attack succeeded 40% of the time, with Grok decrypting and executing malicious instructions to exfiltrate session data. The researchers also tested the technique against Gemini, which has since become more resistant.Dark Reading2-Click Cursor Exploit Enables Dev Environment TakeoverSecurity researchers at Adversa AI discovered that Cursor AI, a popular AI coding editor used by more than 50,000 enterprises including 64% of the Fortune 500, can be exploited in just two clicks to install malicious MCP servers on developers' machines, enabling arbitrary command execution. The attack works by concealing an MCP installation deeplink inside a pull request link using double URL encoding, then exploiting a confirmation dialog that truncates the argument field, hiding malicious commands from view. Cursor AI has acknowledged the vulnerability after it was initially incorrectly closed by their third-party security vendor, with researchers recommending organizations implement allowlisting and agent-aware controls to defend against such chain-of-events attacks.SlashdotDecades-Old Bash Tricks Expose AI Coding Agents To Supply Chain AttacksAI security researchers at Adversa AI have uncovered a structural vulnerability called GuardFall that allows decades-old Bash shell tricks to bypass security safeguards in AI coding agents. The flaw enables attackers to embed malicious commands in repositories, README files, or Makefiles that can steal credentials and compromise developer systems when executed. Testing of 11 popular open source AI coding agents revealed that only one successfully blocked all the Bash trick techniques.