Skip to content

AI security penetration testing providers

42 published entries state AI security penetration testing on their AI testing pages. Between them they name 8 AI product types.

Penetration testing of AI features, including prompt injection, data exfiltration and abuse of agent tools.

prompt injection: Input crafted to override a model's instructions or extract hidden data.
AI product type
Company type
Country (HQ or delivery)

Showing 42 of 42. Default order: coverage score, then name. How the score works.

Entries on this page: AI security penetration testing providers
RankCompany typeAI products (as stated)ServicesVerified
1DeviQAPure-play QA companyPoland, Warsaw
LLM applicationsAI agentsRecommender and ranking systemsPredictive ML modelsComputer visionVoice AI
82010162026-11-09
2ImpactQAPure-play QA companyUnited States
LLM applicationsAI agentsComputer visionVoice AIRecommender and ranking systems
9152026-11-09
4BetterQAPure-play QA companyRomania, Cluj-Napoca
LLM applicationsAI agentsVoice AI
92018142026-11-09
4QATestLabPure-play QA companyCyprus
LLM applicationsAI agentsPredictive ML models
10142026-11-09
4TestingXpertsPure-play QA companyUnited States
LLM applicationsAI agentsVoice AI
10142026-11-09
4TestriqPure-play QA companyIndia, Mira-Bhayandar
LLM applicationsAI agentsPredictive ML models
92010142026-11-09
11Fulminous SoftwareIT services firm with a QA practiceIndia, Jaipur
Voice AIAI agentsAutonomous and embedded AI
72019122026-11-09
14QA MadnessPure-play QA companyPoland, Warsaw
LLM applicationsAI agentsRecommender and ranking systems
7112026-11-09
14QAwerkPure-play QA companyUkraine, Kyiv
LLM applicationsAI agents
72015112026-11-09
18KiwiQAPure-play QA companyAustralia, Bella Vista
LLM applications
72009102026-11-09
18Sparkout TechIT services firm with a QA practiceIndia, Coimbatore
Predictive ML models
72017102026-11-09
23CodoidPure-play QA companyIndia, Chennai
LLM applicationsAI agents
5201292026-11-09
23Luxe QualityPure-play QA companyUnited States, McLean
Computer visionVoice AIPredictive ML models
592026-11-09
29AIDX TechAI assurance and audit specialistSingapore, Singapore
LLM applicationsAI agentsVoice AIComputer vision
482026-11-09
40InnodataData and evaluation vendorUnited States, Ridgefield Park
LLM applicationsGenerative media
462026-11-09
40Pen Test PartnersSecurity firm with AI testingUnited Kingdom, Buckingham
LLM applicationsAI agents
1201062026-11-09
40QualysecSecurity firm with AI testingIndia, Bhubaneswar
LLM applicationsAI agents
362026-11-09
40ScienceSoftIT services firm with a QA practiceUnited States, McKinney
LLM applicationsAI agents
2198962026-11-09
40Software SecuredSecurity firm with AI testingCanada, Ottawa
LLM applicationsAI agents
362026-11-09
50Artifice SecuritySecurity firm with AI testingUnited States, Denver
LLM applicationsAI agents
2201952026-11-09
50Bishop FoxSecurity firm with AI testingUnited States, Tempe
LLM applicationsAI agents
252026-11-09
50BridewellSecurity firm with AI testingUnited Kingdom, Reading
LLM applications
2201352026-11-09
50DeepStrikeSecurity firm with AI testingUnited States
LLM applications
2201652026-11-09
50HackerOneSecurity firm with AI testingUnited States, San Francisco
LLM applicationsAI agents
352026-11-09
50Knovel EngineeringIT services firm with a QA practiceSingapore, Singapore
LLM applications
452026-11-09
50PraetorianSecurity firm with AI testingUnited States, Austin
LLM applicationsAI agents
352026-11-09
50Redbot SecuritySecurity firm with AI testingUnited States, Denver
LLM applicationsAI agents
252026-11-09
617ASecuritySecurity firm with AI testingIreland, Dublin
LLM applicationsAI agents
242026-11-09
61DataArtIT services firm with a QA practiceUnited States, New York
LLM applications
1199742026-11-09
61IBA GroupIT services firm with a QA practiceCzech Republic, Prague
LLM applicationsAI agents
142026-11-09
61IBMIT services firm with a QA practiceUnited States, Armonk
LLM applications
2191142026-11-09
61NexigenIT services firm with a QA practiceUnited States, Newport
LLM applicationsAI agents
242026-11-09
61VaadataSecurity firm with AI testingFrance, Lyon
LLM applicationsAI agents
242026-11-09
61Welo DataData and evaluation vendorUnited States
LLM applications
342026-11-09
61WorkNestUnited Kingdom, London
LLM applicationsAI agents
242026-11-09
74Anvil SecureSecurity firm with AI testingUnited States, Seattle
LLM applications
232026-11-09
74BreachLockSecurity firm with AI testingUnited States, New York
LLM applications
132026-11-09
74BugcrowdSecurity firm with AI testingUnited States, San Francisco
LLM applications
1201232026-11-09
74CrowdStrikeSecurity firm with AI testingUnited States, Austin
LLM applications
232026-11-09
74DionachSecurity firm with AI testingUnited Kingdom, Oxford
LLM applications
232026-11-09
74NetSPISecurity firm with AI testingUnited States, Minneapolis
LLM applications
232026-11-09
86SynackSecurity firm with AI testingUnited States, Redwood City
LLM applications
122026-11-09

Product types covered by vendors stating this service

Count of entries on this page that state each product type.

Product types stated by entries offering AI security penetration testing
Product typeAI security penetration testing
LLM applications39
AI agents25
Voice AI7
Computer vision4
Predictive ML models5
Recommender and ranking systems3
Generative media1
Document AI0
Autonomous and embedded AI1

Scope and standards referenced

Prompt injection, data exfiltration, agent and tool abuse, supply chain; OWASP LLM Top 10.

The directory lists which vendors state this service. It does not test whether a vendor applies these standards.

Questions

What is AI penetration testing?

AI penetration testing attacks an application built on a model the way a real attacker would: through prompts, retrieved documents, tools and integrations. The goal is to find paths to data exposure, unauthorised actions or service abuse. 42 published entries state AI security penetration testing.

What does an LLM penetration test cover?

Typical scope is prompt injection, system prompt and data leakage, insecure output handling, abuse of tools and plugins, access control around retrieved data, and denial of service through costly requests. Many reports map findings to the OWASP Top 10 for LLM Applications.

What is prompt injection testing?

Prompt injection testing tries to override the model's instructions through user input or through content the model reads, such as documents, web pages or emails. Testers check whether injected text can change behaviour, reveal hidden prompts or trigger tool calls.

How do you choose a company to penetration test an AI application?

Check that the vendor's own AI page describes testing of applications like yours, which reference list it maps findings to, whether testers are its own staff, and what the report contains. The table on this page lists every published entry that states this service, ordered by coverage score.

What is MCP security testing?

MCP is the Model Context Protocol, a standard way for AI agents to connect to tools and data sources. MCP security testing checks those connections: tool permissions, authentication, injected instructions in tool output, and what an agent can do through each server.

Does an LLM penetration test check for training data leakage?

Many tests include attempts to extract memorised training data, system prompts or data from connected sources. Whether a given test covers it depends on the scope agreed with the vendor, so ask for it to be named in the statement of work.

Request a listing or correction