- GEN AI SECURITY
- Solution landscape
AI Security Solutions Landscape
The landscape includes traditional and emerging security controls addressing LLM and Generative AI risks in the OWASP Top 10. It is not a comprehensive list or an endorsement but a community resource of open source and proprietary solutions. Contributions are open and reviewed for accuracy.
Watch the video
- Proprietary
- Stage: Develop & Experiment
Pangea Authentication
Pangea
Secure authentication, with support for adaptive threat intelligence, built specifically to protect access to your AI application, protect your users, and your organization.
- LLM04, LLM07, LLM10
- Proprietary
- Stage: Test and Evaluation
Protect AI
Recon runs automated and comprehensive penetration testing attacks on your LLM powered applications, to help protect you from unique security threats and vulnerabilities. It has
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM09
- Proprietary
- Stage: Scope & Plan
Microsoft
Defender for Cloud AI-SPM identifies vulnerabilities and misconfigurations in generative AI apps on Azure OpenAI, Azure Machine Learning, and Amazon Bedrock, providing actionable recommendations and
- LLM04, LLM08
- Proprietary
- Stage: Test & Evaluate
SAIFE X RedTeam
Eroun&Company
RedTeam solution to automate detection of malicious prompt attack vulnerabilities against LLM
- LLM01, LLM02, LLM03, LLM07, LLM10
- Open Source
- Stage: Data Augmentation and Fine-Tuning
Decisionbox
Decisionbox
Decisionbox makes LLM applications learn from data by transforming zero-shot prompts into fine-tuned machine learning classifiers.
- LLM02, LLM03, LLM05
- Proprietary
- Stage: Test and Evaluation
Citadel Lens
Citadel AI
Citadel Lens is a tool for multilingual, automated red teaming and evaluation of LLM applications.
- LLM01, LLM02, LLM06
- Proprietary
- Stage: Test and Evaluation
AIFT
Vulcan is an LLM risk and vulnerability testing solution that enables AI project teams to perform automatic red teaming at scale.
- LLM01, LLM02, LLM04, LLM06, LLM08, LLM09
- Proprietary
- Stage: Monitor
Brand Engagement Networks
Red Teaming / Security Testing in the AI CI/CD. The SPLX.ai platform provides continuous testing, guard rail assessments, domain specific test scenarios, AI Inventory which
- Proprietary
- Stage: Monitor, Test & Evaluate
KELA
AiFort by KELA is an automated, intelligence-led red teaming platform designed to protect GenAI applications. AiFort allows organizations full protection through test simulations of their
- LLM01, LLM02, LLM04, LLM05, LLM06, LLM08, LLM09
- Open Source
- Stage: Test and Evaluation
AIShield,Powered by Bosch
AIShield Watchtower automates model and notebook discovery, performing thorough vulnerability scans to identify risks like hard-coded secrets, PII exposure, outdated libraries, serialization attacks, and unsafe
- LLM03, LLM05, LLM06
- 196
- 14
- Open Source
- Stage: Deploy, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Noma Security
- Adversarial Attack Protection, Adversarial Input Detection, Adversarial Testing, AI/ML Bill of Materials (BOM), Automated Vulnerability Scanning, Code Analysis, Data Privacy and Protection, GenAI/LLM Posture Management, Incident Simulation, LLM & App Vulnerability Scanning, LLM Enabled Web Application Firewall, LLM Guardrails, Model and Application Interaction Security, Penetration Testing, Prompt Security, Response Testing, Runtime Application Self-Protection, SAST/DAST/IAST, Secure CI/CD pipeline, Secure Coding Practices, Secure Library/Code Repository, Secure Output Handling, Software Composition Analysis, Supply Chain Verification, Third-Party Risk Assessment, Vulnerability Scanning
Noma Security
"Noma Security is a comprehensive application security solution for the Data and AI lifecycle. It offers , End-to-End Visibility: Scanning notebooks, source code, and other
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Develop & Experiment, Scope & Plan
SpiceDB
Open source, Google Zanzibar-inspired permissions database for scalably storing and querying fine-grained authorization data.
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM10
- 5300
- 286
- Proprietary
- Stage: Deploy, Govern, Monitor, Operate, Release
Palo Alto Networks AI Runtime Security
- Adversarial Input Detection, AI/ML Bill of Materials (BOM), Compliance Management, Compliance Verification, Data Privacy and Protection, Incident Governance, Network Security Validation, Observability, Regulatory Compliance Tracking, Risk Assessment and Management, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
Palo Alto Networks
Palo Alto Networks AI Runtime Security provides continuous discovery, protection, and monitoring for genAI applications, preventing security risks such as prompt injections, sensitive data leakage,
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Test & Evaluate
Modelscan
modelscan
ModelScan is an open source project from Protect AI that scans models to determine if they contain unsafe code.
- LLM01
- 291
- 60
- Proprietary
- Stage: Govern, Monitor
Infotect Security
IWS scans outbound response traffic in real time for undesirable content and confidential data at layer 4. It is a paradigm shift in web security,
- LLM01, LLM02, LLM05, LLM06, LLM07
- Proprietary
- Stage: Deploy, Operate
Cisco Systems, Inc.
Cisco AI Runtime secures GenAI apps to address threats like prompt injections, sensitive data loss, and compliance concerns. Deploy guardrails around safety, privacy, relevancy, and
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Monitor
AISheild,Powered by Bosch
AIShield Guardian functions as an AI firewall and guardrail, providing secure access control, sensitive data protection, and live monitoring. It safeguards interactions between applications and
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM10
- Open Source
- Stage: Deploy, Operate
IronCore Labs Cloaked AI
IronCore Labs
Encrypts vector embeddings stored in databases while still allowing kNN/aNN searches and preventing vector inversion attacks.
- LLM06
- 10
- 3
- Proprietary
- Stage: Deploy, Monitor, Operate
TrojAI Defend
- Adversarial Attack Protection, Adversarial Input Detection, Compliance Verification, Data Privacy and Protection, LLM Enabled Web Application Firewall, LLM Guardrails, LLM Incident Detectio, Model Behavior Analysis, Regulatory Compliance Tracking, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
TrojAI
TrojAI Defend protects AI models from evolving threats at runtime, including prompt injection, jailbreaking, DoS attacks, data leakage and loss, and toxic or offensive content.
- LLM01, LLM02, LLM04, LLM06, LLM10
- Open Source
- Stage: Test & Evaluate
LLM Vulnerability Scanner
Garak.ai
Garak helps you discover weaknesses and unwanted behaviors in anything using language model technology. With garak, you can scan a chatbot or model and quickly
- LLM01
- 1200
- 141
- Proprietary
- Stage: Operate
Microsoft Security – Secure and Govern AI
Microsoft
Microsoft Security provides capabilities to discover, protect, and govern AI applications. Data Security, AI Security Posture Management, AI Threat Protection, AI governance and more.
- LLM01, LLM02, LLM04, LLM05, LLM06, LLM08, LLM09
- Open Source
- Stage: Deploy, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Noma Security
- Adversarial Attack Protection, Adversarial Input Detection, Adversarial Testing, AI/ML Bill of Materials (BOM), Automated Vulnerability Scanning, Code Analysis, Data Privacy and Protection, GenAI/LLM Posture Management, Incident Simulation, LLM & App Vulnerability Scanning, LLM Enabled Web Application Firewall, LLM Guardrails, Model and Application Interaction Security, Penetration Testing, Prompt Security, Response Testing, Runtime Application Self-Protection, SAST/DAST/IAST, Secure CI/CD pipeline, Secure Coding Practices, Secure Library/Code Repository, Secure Output Handling, Software Composition Analysis, Supply Chain Verification, Third-Party Risk Assessment, Vulnerability Scanning
Noma Security
"Noma Security is a comprehensive application security solution for the Data and AI lifecycle. It offers , End-to-End Visibility: Scanning notebooks, source code, and other
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Develop & Experiment, Scope & Plan
SpiceDB
Open source, Google Zanzibar-inspired permissions database for scalably storing and querying fine-grained authorization data.
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM10
- 5300
- 286
- Proprietary
- Stage: Scope & Plan
Microsoft
Defender for Cloud AI-SPM identifies vulnerabilities and misconfigurations in generative AI apps on Azure OpenAI, Azure Machine Learning, and Amazon Bedrock, providing actionable recommendations and
- LLM04, LLM08
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Deploy, Develop & Experiment, Operate, Release, Scope & Plan, Test & Evaluate
Prisma Cloud AI-SPM
Palo Alto Networks
Prisma Cloud AI-SPM helps organizations discover, classify, protect and govern AI-powered applications. It provides visibility into the entire AI ecosystem including model, applications and resources,
- Proprietary
- Stage: Scope & Plan
Seezo Security Design Review
Seezo.io
Seezo leverages LLMs to provide context-specific security requirements to developers before they start coding
- LLM01, LLM02, LLM07
- Open Source
- Stage: Scope & Plan
StrideGPT
Stride GPT
A threat model helps identify and evaluate potential security threats to applications / systems. It provides a systematic approach to understanding possible vulnerabilities and attack
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Scope & Plan
Mitre ATLAS
Mitre
ATLAS (Adversarial Threat Landscape for Artificial-Intelligence Systems) is a globally accessible, living knowledge base of adversary tactics and techniques against Al-enabled systems based on real-world
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Pillar Security
- Adversarial Attack Protection, Adversarial Testing, AI/ML Bill of Materials (BOM), Data Leakage Protection, Early Identification of Sensitive Data, Ethical Compliance, Incident Simulation, LLM & App Vulnerability Scanning, LLM Benchmarking, LLM Guardrails, Model and Application Interaction Security, Model Behavior Analysis, Penetration Testing, Privacy, Regulatory Compliance Tracking, Response Testing, Runtime Application Self-Protection, Security Alerting, Software Composition Analysis, User Activity Monitoring
Pillar Security
Pillar enables teams to rapidly adopt AI with minimal risk by providing a unified AI security layer across the organization
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08
- Proprietary
- Stage: Augment/Fine Tune Data, Deploy, Develop & Experiment, Govern, Monitor, Operate, Scope & Plan
Securiti
Securiti Data Command Center provides unified intelligence, controls, and orchestration for enabling the safe use of data and AI across hybrid multi-clouds. Enterprises rely on
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Augment/Fine Tune Data
Unstructured.io
Unstructured.io
Unstructured is the leading provider of LLM data preprocessing solutions, empowering organizations to transform their internal unstructured data into formats compatible with large language models
- LLM06
- Proprietary
- Stage: Augment/Fine Tune Data, Deploy, Develop & Experiment, Govern, Monitor, Operate, Scope & Plan
Securiti
Securiti Data Command Center provides unified intelligence, controls, and orchestration for enabling the safe use of data and AI across hybrid multi-clouds. Enterprises rely on
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Develop & Experiment, Monitor, Test & Evaluate
AIandMe
- Adversarial Input Detection, Adversarial Testing, Ethical Compliance, Incident Simulation, LLM & App Vulnerability Scanning, Model and Application Interaction Security, Observability, Penetration Testing, Regulatory Compliance Tracking, Response Testing, Secure Coding Practices, Security Alerting, User Activity Monitoring, Vulnerability Scanning
AIandMe
AIandMe provides an end-to-end platform for testing, securing, and monitoring LLM-based AI systems—combining automated adversarial testing, real-time protection, and human-in-the-loop audits to ensure reliable, compliant,
- LLM01, LLM02, LLM04, LLM07, LLM10
- Open Source
- Stage: Deploy, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Noma Security
- Adversarial Attack Protection, Adversarial Input Detection, Adversarial Testing, AI/ML Bill of Materials (BOM), Automated Vulnerability Scanning, Code Analysis, Data Privacy and Protection, GenAI/LLM Posture Management, Incident Simulation, LLM & App Vulnerability Scanning, LLM Enabled Web Application Firewall, LLM Guardrails, Model and Application Interaction Security, Penetration Testing, Prompt Security, Response Testing, Runtime Application Self-Protection, SAST/DAST/IAST, Secure CI/CD pipeline, Secure Coding Practices, Secure Library/Code Repository, Secure Output Handling, Software Composition Analysis, Supply Chain Verification, Third-Party Risk Assessment, Vulnerability Scanning
Noma Security
"Noma Security is a comprehensive application security solution for the Data and AI lifecycle. It offers , End-to-End Visibility: Scanning notebooks, source code, and other
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Develop & Experiment, Scope & Plan
SpiceDB
Open source, Google Zanzibar-inspired permissions database for scalably storing and querying fine-grained authorization data.
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM10
- 5300
- 286
- Open Source
- Stage: Deploy, Develop & Experiment, Test & Evaluate
Infosys Responsible AI Toolkit
- Adversarial Testing, Application Security Orchestration and Correlation, Bias and Fairness Testing, Final Security Audit, Incident Simulation, LLM & App Vulnerability Scanning, LLM Benchmarking, LLM Enabled Web Application Firewall, Model and Application Interaction Security, Penetration Testing, User and Data Privacy Protections
Infosys
The Infosys Responsible AI Toolkit (Technical Guardrail) is API Based solution designed to ensure the ethical and responsible development of AI applications. By integrating safety,
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- 5
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Deploy, Develop & Experiment, Operate, Release, Scope & Plan, Test & Evaluate
Prisma Cloud AI-SPM
Palo Alto Networks
Prisma Cloud AI-SPM helps organizations discover, classify, protect and govern AI-powered applications. It provides visibility into the entire AI ecosystem including model, applications and resources,
- Proprietary
- Stage: Develop & Experiment, Test & Evaluate
TrojAI
TrojAI Detect secures AI behavior at build time. The AI security platform continuously red teams AI models to find security weaknesses in AI, ML, and
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM09, LLM10
- Proprietary
- Stage: Deploy, Develop & Experiment, Monitor
Operant 3D Runtime Defense
Operant AI
Operant provides runtime application defense with threat detection and remediation, automated policy enforcement, and in-line PII redaction. It secures cloud-native environments, protecting APIs, data flows,
- LLM01, LLM02, LLM04, LLM05, LLM06, LLM07, LLM08, LLM10
- Proprietary
- Stage: Develop & Experiment
Pangea Sanitize
Pangea
Utilize Pangea's Sanitize service to ensure that malicious scripts, malicious links, profanity, and regulated PII are not submitted in prompt inputs, prompt responses, or in
- LLM02, LLM03, LLM05, LLM06
- Proprietary
- Stage: Develop & Experiment
Pangea
Pangea's Authorization service is an access control engine that integrates with any AI application through easy-to-use APIs and SDKs. It is used to enforce access
- LLM04, LLM06, LLM07, LLM08, LLM10
- Proprietary
- Stage: Develop & Experiment
Pangea Authentication
Pangea
Secure authentication, with support for adaptive threat intelligence, built specifically to protect access to your AI application, protect your users, and your organization.
- LLM04, LLM07, LLM10
- Proprietary
- Stage: Develop & Experiment
Pangea
Protect your users and application by redacting sensitive info from prompt inputs, prompt responses, and contextual data, using Pangea's Redact service.
- LLM01, LLM02, LLM03, LLM06
- Proprietary
- Stage: Develop & Experiment
Pangea Data Guard
Pangea
Prompt inputs, responses, and data ingestion from external sources can all be evaluated for malicious content with Pangea's Data Guard to protect LLMs and users
- LLM02, LLM03, LLM07, LLM10
- Open Source
- Stage: Deploy, Develop & Experiment
PurpleLlama CodeShield
Meta
CodeShield is an effort to mitigate against the insecure code generated by LLMs. CodeShield is a robust inference time filtering tool engineered to prevent the
- LLM02
- Proprietary
- Stage: Develop & Experiment
Pangea Prompt Guard
Pangea
Pangea's Prompt Guard service utilizes a deep understanding of prompt templates, heuristics and trained models to detect direct or indirect prompt injection attacks and jailbreak
- LLM01, LLM03
- Proprietary
- Stage: Develop & Experiment, Govern, Monitor, Release, Test & Evaluate
Cisco Systems
Cisco AI Validation assesses AI applications and models for security and safety vulnerabilities. We automatically analyze a model’s risk across hundreds of attack techniques and
- LLM01, LLM03, LLM04, LLM06, LLM09
- Proprietary
- Stage: Develop & Experiment
Mend AI
Mend AI provides a shift-left solution for securing AI-driven applications. It enables discovery of shadow AI, security and compliance analysis through code scanning and red-teaming,
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Deploy, Develop & Experiment, Monitor, Test & Evaluate
Aqua Security
Aqua facilitates secure application development and runtime protection by addressing vulnerabilities outlined in the OWASP Top 10 for LLM applications.
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Develop & Experiment
Flickling
Trail of Bits
Fickling can help securing AI/ML codebases by automatically scanning pickle files contained in models. Fickling hooks the pickle module and verifies imports made when loading
- LLM03
- 398
- 43
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Pillar Security
- Adversarial Attack Protection, Adversarial Testing, AI/ML Bill of Materials (BOM), Data Leakage Protection, Early Identification of Sensitive Data, Ethical Compliance, Incident Simulation, LLM & App Vulnerability Scanning, LLM Benchmarking, LLM Guardrails, Model and Application Interaction Security, Model Behavior Analysis, Penetration Testing, Privacy, Regulatory Compliance Tracking, Response Testing, Runtime Application Self-Protection, Security Alerting, Software Composition Analysis, User Activity Monitoring
Pillar Security
Pillar enables teams to rapidly adopt AI with minimal risk by providing a unified AI security layer across the organization
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08
- Proprietary
- Stage: Augment/Fine Tune Data, Deploy, Develop & Experiment, Govern, Monitor, Operate, Scope & Plan
Securiti
Securiti Data Command Center provides unified intelligence, controls, and orchestration for enabling the safe use of data and AI across hybrid multi-clouds. Enterprises rely on
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Govern, Monitor, Release, Test & Evaluate
The CalypsoAI Inference Platform
- Adversarial Testing, AI/LLM Secure Posture Management, AI/ML Bill of Materials (BOM), Application Security Orchestration and Correlation, Bias and Fairness Oversight, Bias and Fairness Testing, Compliance Management, Data Privacy and Protection, Data Security Posture Management, Digital Model Signing, Ethical Compliance, Final Security Audit, Incident Governance, Incident Simulation, LLM Benchmarking, Model Behavior Analysis, Model Security Posture Evaluation, Observability, Patch and Update Alerts, Penetration Testing, Regulatory Compliance Tracking, Response Testing, Risk Assessment and Management, SAST/DAST/IAST, Secure CI/CD pipeline, Security Alerting, Security Metrics Collection, Supply Chain Verification, User Access Control Validation, User Activity Monitoring, User/Machine Access audits, Vulnerability Scanning
CalypsoAI
CalypsoAI secures GenAI across applications and agents. The CalypsoAI Inference Platform tests, defends, and monitors AI in development and production. With Defend, Red-Team, and Observe,
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM10
- Proprietary
- Stage: Develop & Experiment, Monitor, Test & Evaluate
AIandMe
- Adversarial Input Detection, Adversarial Testing, Ethical Compliance, Incident Simulation, LLM & App Vulnerability Scanning, Model and Application Interaction Security, Observability, Penetration Testing, Regulatory Compliance Tracking, Response Testing, Secure Coding Practices, Security Alerting, User Activity Monitoring, Vulnerability Scanning
AIandMe
AIandMe provides an end-to-end platform for testing, securing, and monitoring LLM-based AI systems—combining automated adversarial testing, real-time protection, and human-in-the-loop audits to ensure reliable, compliant,
- LLM01, LLM02, LLM04, LLM07, LLM10
- Proprietary
- Stage: Test & Evaluate
SAIFE X RedTeam
Eroun&Company
RedTeam solution to automate detection of malicious prompt attack vulnerabilities against LLM
- LLM01, LLM02, LLM03, LLM07, LLM10
- Open Source
- Stage: Deploy, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Noma Security
- Adversarial Attack Protection, Adversarial Input Detection, Adversarial Testing, AI/ML Bill of Materials (BOM), Automated Vulnerability Scanning, Code Analysis, Data Privacy and Protection, GenAI/LLM Posture Management, Incident Simulation, LLM & App Vulnerability Scanning, LLM Enabled Web Application Firewall, LLM Guardrails, Model and Application Interaction Security, Penetration Testing, Prompt Security, Response Testing, Runtime Application Self-Protection, SAST/DAST/IAST, Secure CI/CD pipeline, Secure Coding Practices, Secure Library/Code Repository, Secure Output Handling, Software Composition Analysis, Supply Chain Verification, Third-Party Risk Assessment, Vulnerability Scanning
Noma Security
"Noma Security is a comprehensive application security solution for the Data and AI lifecycle. It offers , End-to-End Visibility: Scanning notebooks, source code, and other
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Monitor, Test & Evaluate
KELA
AiFort by KELA is an automated, intelligence-led red teaming platform designed to protect GenAI applications. AiFort allows organizations full protection through test simulations of their
- LLM01, LLM02, LLM04, LLM05, LLM06, LLM08, LLM09
- Proprietary
- Stage: Monitor, Operate, Test & Evaluate
Straiker AI
- Adversarial Attack Protection, Adversarial Testing, AI/LLM Secure Posture Management, Application Security Orchestration and Correlation, Bias and Fairness Testing, Data Integrity and Encryption, Data Leakage Protection, LLM Benchmarking, LLM Guardrails, LLM Incident Detection and Response, Observability, Penetration Testing, Prompt Security, Secure Output Handling, Security Alerting, Security Metrics Collection, User Activity Monitoring
Straiker Inc
Secure AI Applications using two products. Ascend AI provides pentesting/red teaming across all layers of the applications. Defend AI provides visibility, guardrails for AI applications.
- LLM01, LLM02, LLM05, LLM06, LLM07, LLM09
- Proprietary
- Stage: Test & Evaluate
AIM Intelligence
AIM Supervisor integrates AIM RED for automated AI vulnerability testing, AIM GUARD for real-time threat detection and mitigation, and AIM Benchmark for comprehensive safety evaluations,
- LLM01, LLM02, LLM03, LLM04
- Proprietary
- Stage: Test & Evaluate
Adversa AI Red Teaming Platform
Adversa AI
Adversa AI's Red Teaming platform provides automated security testing of Generative AI systems, identifying all possible vulnerabilities like jailbreaks, prompt injections, and adversarial attacks to
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Deploy, Develop & Experiment, Test & Evaluate
Infosys Responsible AI Toolkit
- Adversarial Testing, Application Security Orchestration and Correlation, Bias and Fairness Testing, Final Security Audit, Incident Simulation, LLM & App Vulnerability Scanning, LLM Benchmarking, LLM Enabled Web Application Firewall, Model and Application Interaction Security, Penetration Testing, User and Data Privacy Protections
Infosys
The Infosys Responsible AI Toolkit (Technical Guardrail) is API Based solution designed to ensure the ethical and responsible development of AI applications. By integrating safety,
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- 5
- Proprietary
- Stage: Govern, Monitor, Operate, Test & Evaluate
DynamoGuard
- Adversarial Attack Protection, Adversarial Testing, Compliance Management, Data Leakage Protection, Ethical Compliance, LLM Benchmarking, LLM Guardrails, Observability, Penetration Testing, Prompt Security, Regulatory Compliance Tracking, Risk Assessment and Management, User Activity Monitoring, Vulnerability Scanning
Dynamo AI
DynamoGuard offers real-time guardrailing for GenAI, customizable in natural language and capable of running in the cloud, hybrid, on-prem or fully on edge devices to
- LLM01, LLM06, LLM09
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Deploy, Develop & Experiment, Operate, Release, Scope & Plan, Test & Evaluate
Prisma Cloud AI-SPM
Palo Alto Networks
Prisma Cloud AI-SPM helps organizations discover, classify, protect and govern AI-powered applications. It provides visibility into the entire AI ecosystem including model, applications and resources,
- Proprietary
- Stage: Develop & Experiment, Test & Evaluate
TrojAI
TrojAI Detect secures AI behavior at build time. The AI security platform continuously red teams AI models to find security weaknesses in AI, ML, and
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM09, LLM10
- Open Source
- Stage: Test & Evaluate
Modelscan
modelscan
ModelScan is an open source project from Protect AI that scans models to determine if they contain unsafe code.
- LLM01
- 291
- 60
- Open Source
- Stage: Test & Evaluate
CyberSecEval
Meta
CyberSecEval is an extensive benchmark suite under Meta PurpleLlama, designed to evaluate various cybersecurity risks of LLMs, including several listed in the OWASP Top-10 for
- LLM01, LLM02, LLM07, LLM08, LLM09, LLM10
- 2600
- 439
- Proprietary
- Stage: Develop & Experiment, Govern, Monitor, Release, Test & Evaluate
Cisco Systems
Cisco AI Validation assesses AI applications and models for security and safety vulnerabilities. We automatically analyze a model’s risk across hundreds of attack techniques and
- LLM01, LLM03, LLM04, LLM06, LLM09
- Proprietary
- Stage: Test & Evaluate
Enkrypt AI
Enkrypt AI secures enterprises against generative AI risks with its comprehensive security platform that detects threats, removes vulnerabilities, and monitors the latest insights on security,
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Deploy, Develop & Experiment, Monitor, Test & Evaluate
Aqua Security
Aqua facilitates secure application development and runtime protection by addressing vulnerabilities outlined in the OWASP Top 10 for LLM applications.
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Test & Evaluate
Harmbench
HarmBench is a new evaluation framework for automated red teaming and robust refusal.
- LLM01, LLM02, LLM03, LLM06, LLM08, LLM09
- 290
- 49
- Open Source
- Stage: Test & Evaluate
Prompt Fuzzer
Prompt Security
Prompt Fuzzer is an interactive, open-source tool that empowers developers of GenAI applications to evaluate and enhance the resilience and safety of their system prompts.
- LLM01, LLM02, LLM03, LLM06
- 361
- 44
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Pillar Security
- Adversarial Attack Protection, Adversarial Testing, AI/ML Bill of Materials (BOM), Data Leakage Protection, Early Identification of Sensitive Data, Ethical Compliance, Incident Simulation, LLM & App Vulnerability Scanning, LLM Benchmarking, LLM Guardrails, Model and Application Interaction Security, Model Behavior Analysis, Penetration Testing, Privacy, Regulatory Compliance Tracking, Response Testing, Runtime Application Self-Protection, Security Alerting, Software Composition Analysis, User Activity Monitoring
Pillar Security
Pillar enables teams to rapidly adopt AI with minimal risk by providing a unified AI security layer across the organization
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08
- Open Source
- Stage: Deploy, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Noma Security
- Adversarial Attack Protection, Adversarial Input Detection, Adversarial Testing, AI/ML Bill of Materials (BOM), Automated Vulnerability Scanning, Code Analysis, Data Privacy and Protection, GenAI/LLM Posture Management, Incident Simulation, LLM & App Vulnerability Scanning, LLM Enabled Web Application Firewall, LLM Guardrails, Model and Application Interaction Security, Penetration Testing, Prompt Security, Response Testing, Runtime Application Self-Protection, SAST/DAST/IAST, Secure CI/CD pipeline, Secure Coding Practices, Secure Library/Code Repository, Secure Output Handling, Software Composition Analysis, Supply Chain Verification, Third-Party Risk Assessment, Vulnerability Scanning
Noma Security
"Noma Security is a comprehensive application security solution for the Data and AI lifecycle. It offers , End-to-End Visibility: Scanning notebooks, source code, and other
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Deploy, Monitor, Operate
AI Trust Platform
- Adversarial Attack Protection, Adversarial Input Detection, AI/LLM Secure Posture Management, Data Leakage Protection, Data Privacy and Protection, Ethical Compliance, LLM Guardrails, LLM Incident Detection and Response, Observability, Privacy, Prompt Security, Runtime Application Self-Protection, Secure Output Handling, Security Alerting, Security Metrics Collection, User Activity Monitoring, User and Data Privacy Protections
Preamble
Preamble provides runtime guardrails for RAG, LLMs, and AI agents by enforcing safety, privacy, security, and compliance policies while mitigating real-time risks to ensure secure,
- LLM01, LLM02, LLM03, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Deploy, Develop & Experiment, Test & Evaluate
Infosys Responsible AI Toolkit
- Adversarial Testing, Application Security Orchestration and Correlation, Bias and Fairness Testing, Final Security Audit, Incident Simulation, LLM & App Vulnerability Scanning, LLM Benchmarking, LLM Enabled Web Application Firewall, Model and Application Interaction Security, Penetration Testing, User and Data Privacy Protections
Infosys
The Infosys Responsible AI Toolkit (Technical Guardrail) is API Based solution designed to ensure the ethical and responsible development of AI applications. By integrating safety,
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- 5
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Deploy, Develop & Experiment, Operate, Release, Scope & Plan, Test & Evaluate
Prisma Cloud AI-SPM
Palo Alto Networks
Prisma Cloud AI-SPM helps organizations discover, classify, protect and govern AI-powered applications. It provides visibility into the entire AI ecosystem including model, applications and resources,
- Proprietary
- Stage: Deploy, Monitor, Operate
TrojAI Defend
- Adversarial Attack Protection, Adversarial Input Detection, Compliance Verification, Data Privacy and Protection, LLM Enabled Web Application Firewall, LLM Guardrails, LLM Incident Detectio, Model Behavior Analysis, Regulatory Compliance Tracking, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
TrojAI
TrojAI Defend protects AI models from evolving threats at runtime, including prompt injection, jailbreaking, DoS attacks, data leakage and loss, and toxic or offensive content.
- LLM01, LLM02, LLM04, LLM06, LLM10
- Proprietary
- Stage: Deploy, Develop & Experiment, Monitor
Operant 3D Runtime Defense
Operant AI
Operant provides runtime application defense with threat detection and remediation, automated policy enforcement, and in-line PII redaction. It secures cloud-native environments, protecting APIs, data flows,
- LLM01, LLM02, LLM04, LLM05, LLM06, LLM07, LLM08, LLM10
- Proprietary
- Stage: Deploy, Govern, Monitor, Operate, Release
Palo Alto Networks AI Runtime Security
- Adversarial Input Detection, AI/ML Bill of Materials (BOM), Compliance Management, Compliance Verification, Data Privacy and Protection, Incident Governance, Network Security Validation, Observability, Regulatory Compliance Tracking, Risk Assessment and Management, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
Palo Alto Networks
Palo Alto Networks AI Runtime Security provides continuous discovery, protection, and monitoring for genAI applications, preventing security risks such as prompt injections, sensitive data leakage,
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Deploy, Develop & Experiment
PurpleLlama CodeShield
Meta
CodeShield is an effort to mitigate against the insecure code generated by LLMs. CodeShield is a robust inference time filtering tool engineered to prevent the
- LLM02
- Proprietary
- Stage: Deploy, Operate
Cisco Systems, Inc.
Cisco AI Runtime secures GenAI apps to address threats like prompt injections, sensitive data loss, and compliance concerns. Deploy guardrails around safety, privacy, relevancy, and
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Deploy, Develop & Experiment, Monitor, Test & Evaluate
Aqua Security
Aqua facilitates secure application development and runtime protection by addressing vulnerabilities outlined in the OWASP Top 10 for LLM applications.
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Deploy, Operate
IronCore Labs Cloaked AI
IronCore Labs
Encrypts vector embeddings stored in databases while still allowing kNN/aNN searches and preventing vector inversion attacks.
- LLM06
- 10
- 3
- Proprietary
- Stage: Augment/Fine Tune Data, Deploy, Develop & Experiment, Govern, Monitor, Operate, Scope & Plan
Securiti
Securiti Data Command Center provides unified intelligence, controls, and orchestration for enabling the safe use of data and AI across hybrid multi-clouds. Enterprises rely on
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Operate
Insight For Webserver (IWS)
Infotect Security
IWS scans outbound response traffic in real time for undesirable content and confidential data at layer 4. It is a paradigm shift in web security.
- LLM01, LLM02, LLM05, LLM06, LLM07
- Open Source
- Stage: Deploy, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Noma Security
- Adversarial Attack Protection, Adversarial Input Detection, Adversarial Testing, AI/ML Bill of Materials (BOM), Automated Vulnerability Scanning, Code Analysis, Data Privacy and Protection, GenAI/LLM Posture Management, Incident Simulation, LLM & App Vulnerability Scanning, LLM Enabled Web Application Firewall, LLM Guardrails, Model and Application Interaction Security, Penetration Testing, Prompt Security, Response Testing, Runtime Application Self-Protection, SAST/DAST/IAST, Secure CI/CD pipeline, Secure Coding Practices, Secure Library/Code Repository, Secure Output Handling, Software Composition Analysis, Supply Chain Verification, Third-Party Risk Assessment, Vulnerability Scanning
Noma Security
"Noma Security is a comprehensive application security solution for the Data and AI lifecycle. It offers , End-to-End Visibility: Scanning notebooks, source code, and other
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Operate
Microsoft Security – Secure and Govern AI
Microsoft
Microsoft Security provides capabilities to discover, protect, and govern AI applications. Data Security, AI Security Posture Management, AI Threat Protection, AI governance and more.
- LLM01, LLM02, LLM04, LLM05, LLM06, LLM08, LLM09
- Proprietary
- Stage: Deploy, Monitor, Operate
AI Trust Platform
- Adversarial Attack Protection, Adversarial Input Detection, AI/LLM Secure Posture Management, Data Leakage Protection, Data Privacy and Protection, Ethical Compliance, LLM Guardrails, LLM Incident Detection and Response, Observability, Privacy, Prompt Security, Runtime Application Self-Protection, Secure Output Handling, Security Alerting, Security Metrics Collection, User Activity Monitoring, User and Data Privacy Protections
Preamble
Preamble provides runtime guardrails for RAG, LLMs, and AI agents by enforcing safety, privacy, security, and compliance policies while mitigating real-time risks to ensure secure,
- LLM01, LLM02, LLM03, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Operate
WebOrion® Protector Plus
Cloudsine Pte Ltd
WebOrion® Protector Plus is a GenAI firewall, built to protect GenAI applications against cyber threats. Its ShieldPrompt™ add-on offers an advanced level of protection, including
- LLM01, LLM02, LLM05, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Monitor, Operate, Test & Evaluate
Straiker AI
- Adversarial Attack Protection, Adversarial Testing, AI/LLM Secure Posture Management, Application Security Orchestration and Correlation, Bias and Fairness Testing, Data Integrity and Encryption, Data Leakage Protection, LLM Benchmarking, LLM Guardrails, LLM Incident Detection and Response, Observability, Penetration Testing, Prompt Security, Secure Output Handling, Security Alerting, Security Metrics Collection, User Activity Monitoring
Straiker Inc
Secure AI Applications using two products. Ascend AI provides pentesting/red teaming across all layers of the applications. Defend AI provides visibility, guardrails for AI applications.
- LLM01, LLM02, LLM05, LLM06, LLM07, LLM09
- Open Source
- Stage: Operate
Dyana
Dyana is a sandbox environment using Docker and Tracee for loading, running and profiling a wide range of files, including machine learning models, ELF executables,
- LLM03, LLM04
- 109
- 4
- Proprietary
- Stage: Operate
F5 AI Gateway
F5
F5 AI Gateway is an advanced security solution that protects, accelerates, and observes AI-powered applications.
- LLM01, LLM02, LLM05, LLM10
- Proprietary
- Stage: Govern, Monitor, Operate, Test & Evaluate
DynamoGuard
- Adversarial Attack Protection, Adversarial Testing, Compliance Management, Data Leakage Protection, Ethical Compliance, LLM Benchmarking, LLM Guardrails, Observability, Penetration Testing, Prompt Security, Regulatory Compliance Tracking, Risk Assessment and Management, User Activity Monitoring, Vulnerability Scanning
Dynamo AI
DynamoGuard offers real-time guardrailing for GenAI, customizable in natural language and capable of running in the cloud, hybrid, on-prem or fully on edge devices to
- LLM01, LLM06, LLM09
- Proprietary
- Stage: Operate
Knostic
Knostic identifies data leakage from LLM-powered enterprise search and provides need-to-know based access controls, ensuring employees receive only the information necessary for their roles, thereby
- LLM02, LLM05, LLM06
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Deploy, Develop & Experiment, Operate, Release, Scope & Plan, Test & Evaluate
Prisma Cloud AI-SPM
Palo Alto Networks
Prisma Cloud AI-SPM helps organizations discover, classify, protect and govern AI-powered applications. It provides visibility into the entire AI ecosystem including model, applications and resources,
- Proprietary
- Stage: Deploy, Monitor, Operate
TrojAI Defend
- Adversarial Attack Protection, Adversarial Input Detection, Compliance Verification, Data Privacy and Protection, LLM Enabled Web Application Firewall, LLM Guardrails, LLM Incident Detectio, Model Behavior Analysis, Regulatory Compliance Tracking, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
TrojAI
TrojAI Defend protects AI models from evolving threats at runtime, including prompt injection, jailbreaking, DoS attacks, data leakage and loss, and toxic or offensive content.
- LLM01, LLM02, LLM04, LLM06, LLM10
- Proprietary
- Stage: Deploy, Govern, Monitor, Operate, Release
Palo Alto Networks AI Runtime Security
- Adversarial Input Detection, AI/ML Bill of Materials (BOM), Compliance Management, Compliance Verification, Data Privacy and Protection, Incident Governance, Network Security Validation, Observability, Regulatory Compliance Tracking, Risk Assessment and Management, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
Palo Alto Networks
Palo Alto Networks AI Runtime Security provides continuous discovery, protection, and monitoring for genAI applications, preventing security risks such as prompt injections, sensitive data leakage,
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Govern, Monitor, Operate
Blueteam AI Gateway
- Adversarial Attack Protection, Bias and Fairness Oversight, Compliance Management, Data Integrity and Encryption, Data Leakage Protection, Data Security Posture Management, LLM Guardrails, Privacy, Prompt Security, Runtime Application Self-Protection, Secure Output Handling, User/Machine Access audits
Blueteam AI
Blueteam AI Gateway is a network-layer appliance that intercepts traffic to AI models and discovers AI use, safeguards data from leaking, and governs safe and
- LLM01, LLM04, LLM06, LLM09
- Proprietary
- Stage: Govern, Monitor, Operate
Aim AI Security Platform
- Adversarial Attack Protection, Automated Vulnerability Scanning, Compliance Management, Data Leakage Protection, Data Security Posture Management, LLM Guardrails, LLM Incident Detection and Response, Privacy, Prompt Security, Risk Assessment and Management, Runtime Application Self-Protection, Secure Output Handling, User/Machine Access audits
Aim Security
The Aim AI Security Platform enables enterprises to secure every AI interaction throughout their AI adoption journey, from AI applications used directly by employees to
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08
- Open Source
- Stage: Operate
Llama Guard
Meta
Llama Guard is a set of LLM system safeguards designed to support developers to detect various common types of violating content across multiple use cases
- LLM01, LLM02, LLM06, LLM07
- 2600
- 440
- Proprietary
- Stage: Deploy, Operate
Cisco Systems, Inc.
Cisco AI Runtime secures GenAI apps to address threats like prompt injections, sensitive data loss, and compliance concerns. Deploy guardrails around safety, privacy, relevancy, and
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Operate
AI Blue Team Service
NRI Secure
AI Blue Team Service provides continuous security monitoring for AI systems, specializing in Large Language Models. It detects AI-specific threats like prompt injection and sensitive
- LLM01, LLM02, LLM04, LLM06, LLM08, LLM09
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Pillar Security
- Adversarial Attack Protection, Adversarial Testing, AI/ML Bill of Materials (BOM), Data Leakage Protection, Early Identification of Sensitive Data, Ethical Compliance, Incident Simulation, LLM & App Vulnerability Scanning, LLM Benchmarking, LLM Guardrails, Model and Application Interaction Security, Model Behavior Analysis, Penetration Testing, Privacy, Regulatory Compliance Tracking, Response Testing, Runtime Application Self-Protection, Security Alerting, Software Composition Analysis, User Activity Monitoring
Pillar Security
Pillar enables teams to rapidly adopt AI with minimal risk by providing a unified AI security layer across the organization
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08
- Proprietary
- Stage: Govern, Monitor, Release, Test & Evaluate
The CalypsoAI Inference Platform
- Adversarial Testing, AI/LLM Secure Posture Management, AI/ML Bill of Materials (BOM), Application Security Orchestration and Correlation, Bias and Fairness Oversight, Bias and Fairness Testing, Compliance Management, Data Privacy and Protection, Data Security Posture Management, Digital Model Signing, Ethical Compliance, Final Security Audit, Incident Governance, Incident Simulation, LLM Benchmarking, Model Behavior Analysis, Model Security Posture Evaluation, Observability, Patch and Update Alerts, Penetration Testing, Regulatory Compliance Tracking, Response Testing, Risk Assessment and Management, SAST/DAST/IAST, Secure CI/CD pipeline, Security Alerting, Security Metrics Collection, Supply Chain Verification, User Access Control Validation, User Activity Monitoring, User/Machine Access audits, Vulnerability Scanning
CalypsoAI
CalypsoAI secures GenAI across applications and agents. The CalypsoAI Inference Platform tests, defends, and monitors AI in development and production. With Defend, Red-Team, and Observe,
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM10
- Proprietary
- Stage: Develop & Experiment, Monitor, Test & Evaluate
AIandMe
- Adversarial Input Detection, Adversarial Testing, Ethical Compliance, Incident Simulation, LLM & App Vulnerability Scanning, Model and Application Interaction Security, Observability, Penetration Testing, Regulatory Compliance Tracking, Response Testing, Secure Coding Practices, Security Alerting, User Activity Monitoring, Vulnerability Scanning
AIandMe
AIandMe provides an end-to-end platform for testing, securing, and monitoring LLM-based AI systems—combining automated adversarial testing, real-time protection, and human-in-the-loop audits to ensure reliable, compliant,
- LLM01, LLM02, LLM04, LLM07, LLM10
- Open Source
- Stage: Deploy, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Noma Security
- Adversarial Attack Protection, Adversarial Input Detection, Adversarial Testing, AI/ML Bill of Materials (BOM), Automated Vulnerability Scanning, Code Analysis, Data Privacy and Protection, GenAI/LLM Posture Management, Incident Simulation, LLM & App Vulnerability Scanning, LLM Enabled Web Application Firewall, LLM Guardrails, Model and Application Interaction Security, Penetration Testing, Prompt Security, Response Testing, Runtime Application Self-Protection, SAST/DAST/IAST, Secure CI/CD pipeline, Secure Coding Practices, Secure Library/Code Repository, Secure Output Handling, Software Composition Analysis, Supply Chain Verification, Third-Party Risk Assessment, Vulnerability Scanning
Noma Security
"Noma Security is a comprehensive application security solution for the Data and AI lifecycle. It offers , End-to-End Visibility: Scanning notebooks, source code, and other
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Monitor, Test & Evaluate
KELA
AiFort by KELA is an automated, intelligence-led red teaming platform designed to protect GenAI applications. AiFort allows organizations full protection through test simulations of their
- LLM01, LLM02, LLM04, LLM05, LLM06, LLM08, LLM09
- Proprietary
- Stage: Deploy, Monitor, Operate
AI Trust Platform
- Adversarial Attack Protection, Adversarial Input Detection, AI/LLM Secure Posture Management, Data Leakage Protection, Data Privacy and Protection, Ethical Compliance, LLM Guardrails, LLM Incident Detection and Response, Observability, Privacy, Prompt Security, Runtime Application Self-Protection, Secure Output Handling, Security Alerting, Security Metrics Collection, User Activity Monitoring, User and Data Privacy Protections
Preamble
Preamble provides runtime guardrails for RAG, LLMs, and AI agents by enforcing safety, privacy, security, and compliance policies while mitigating real-time risks to ensure secure,
- LLM01, LLM02, LLM03, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Monitor, Operate, Test & Evaluate
Straiker AI
- Adversarial Attack Protection, Adversarial Testing, AI/LLM Secure Posture Management, Application Security Orchestration and Correlation, Bias and Fairness Testing, Data Integrity and Encryption, Data Leakage Protection, LLM Benchmarking, LLM Guardrails, LLM Incident Detection and Response, Observability, Penetration Testing, Prompt Security, Secure Output Handling, Security Alerting, Security Metrics Collection, User Activity Monitoring
Straiker Inc
Secure AI Applications using two products. Ascend AI provides pentesting/red teaming across all layers of the applications. Defend AI provides visibility, guardrails for AI applications.
- LLM01, LLM02, LLM05, LLM06, LLM07, LLM09
- Proprietary
- Stage: Govern, Monitor
Infotect Security
IWS scans outbound response traffic in real time for undesirable content and confidential data at layer 4. It is a paradigm shift in web security,
- LLM01, LLM02, LLM05, LLM06, LLM07
- Proprietary
- Stage: Govern, Monitor, Operate, Test & Evaluate
DynamoGuard
- Adversarial Attack Protection, Adversarial Testing, Compliance Management, Data Leakage Protection, Ethical Compliance, LLM Benchmarking, LLM Guardrails, Observability, Penetration Testing, Prompt Security, Regulatory Compliance Tracking, Risk Assessment and Management, User Activity Monitoring, Vulnerability Scanning
Dynamo AI
DynamoGuard offers real-time guardrailing for GenAI, customizable in natural language and capable of running in the cloud, hybrid, on-prem or fully on edge devices to
- LLM01, LLM06, LLM09
- Proprietary
- Stage: Monitor
AISheild,Powered by Bosch
AIShield Guardian functions as an AI firewall and guardrail, providing secure access control, sensitive data protection, and live monitoring. It safeguards interactions between applications and
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08, LLM10
- Proprietary
- Stage: Deploy, Monitor, Operate
TrojAI Defend
- Adversarial Attack Protection, Adversarial Input Detection, Compliance Verification, Data Privacy and Protection, LLM Enabled Web Application Firewall, LLM Guardrails, LLM Incident Detectio, Model Behavior Analysis, Regulatory Compliance Tracking, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
TrojAI
TrojAI Defend protects AI models from evolving threats at runtime, including prompt injection, jailbreaking, DoS attacks, data leakage and loss, and toxic or offensive content.
- LLM01, LLM02, LLM04, LLM06, LLM10
- Proprietary
- Stage: Deploy, Develop & Experiment, Monitor
Operant 3D Runtime Defense
Operant AI
Operant provides runtime application defense with threat detection and remediation, automated policy enforcement, and in-line PII redaction. It secures cloud-native environments, protecting APIs, data flows,
- LLM01, LLM02, LLM04, LLM05, LLM06, LLM07, LLM08, LLM10
- Proprietary
- Stage: Deploy, Govern, Monitor, Operate, Release
Palo Alto Networks AI Runtime Security
- Adversarial Input Detection, AI/ML Bill of Materials (BOM), Compliance Management, Compliance Verification, Data Privacy and Protection, Incident Governance, Network Security Validation, Observability, Regulatory Compliance Tracking, Risk Assessment and Management, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
Palo Alto Networks
Palo Alto Networks AI Runtime Security provides continuous discovery, protection, and monitoring for genAI applications, preventing security risks such as prompt injections, sensitive data leakage,
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Govern, Monitor, Operate
Blueteam AI Gateway
- Adversarial Attack Protection, Bias and Fairness Oversight, Compliance Management, Data Integrity and Encryption, Data Leakage Protection, Data Security Posture Management, LLM Guardrails, Privacy, Prompt Security, Runtime Application Self-Protection, Secure Output Handling, User/Machine Access audits
Blueteam AI
Blueteam AI Gateway is a network-layer appliance that intercepts traffic to AI models and discovers AI use, safeguards data from leaking, and governs safe and
- LLM01, LLM04, LLM06, LLM09
- Proprietary
- Stage: Govern, Monitor, Operate
Aim AI Security Platform
- Adversarial Attack Protection, Automated Vulnerability Scanning, Compliance Management, Data Leakage Protection, Data Security Posture Management, LLM Guardrails, LLM Incident Detection and Response, Privacy, Prompt Security, Risk Assessment and Management, Runtime Application Self-Protection, Secure Output Handling, User/Machine Access audits
Aim Security
The Aim AI Security Platform enables enterprises to secure every AI interaction throughout their AI adoption journey, from AI applications used directly by employees to
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08
- Proprietary
- Stage: Monitor
Protect AI
Enable detection and response across all enterprise LLM applications.
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM07, LLM08, LLM09
- Proprietary
- Stage: Monitor
Lakera
Lakera is an AI Application Firewall that protects against prompt attacks, data loss, and inappropriate content. Lakera integrates with a single line of code and
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Open Source
- Stage: Monitor
PromptGuard
Meta
PromptGuard is a lightweight, low-latency model for detecting prompt injections and jailbreaks. The model sees significant iteration driven by community adoption and feedback, making it
- LLM01
- 440
- Proprietary
- Stage: Develop & Experiment, Govern, Monitor, Release, Test & Evaluate
Cisco Systems
Cisco AI Validation assesses AI applications and models for security and safety vulnerabilities. We automatically analyze a model’s risk across hundreds of attack techniques and
- LLM01, LLM03, LLM04, LLM06, LLM09
- Proprietary
- Stage: Deploy, Develop & Experiment, Monitor, Test & Evaluate
Aqua Security
Aqua facilitates secure application development and runtime protection by addressing vulnerabilities outlined in the OWASP Top 10 for LLM applications.
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Govern, Monitor, Release, Test & Evaluate
The CalypsoAI Inference Platform
- Adversarial Testing, AI/LLM Secure Posture Management, AI/ML Bill of Materials (BOM), Application Security Orchestration and Correlation, Bias and Fairness Oversight, Bias and Fairness Testing, Compliance Management, Data Privacy and Protection, Data Security Posture Management, Digital Model Signing, Ethical Compliance, Final Security Audit, Incident Governance, Incident Simulation, LLM Benchmarking, Model Behavior Analysis, Model Security Posture Evaluation, Observability, Patch and Update Alerts, Penetration Testing, Regulatory Compliance Tracking, Response Testing, Risk Assessment and Management, SAST/DAST/IAST, Secure CI/CD pipeline, Security Alerting, Security Metrics Collection, Supply Chain Verification, User Access Control Validation, User Activity Monitoring, User/Machine Access audits, Vulnerability Scanning
CalypsoAI
CalypsoAI secures GenAI across applications and agents. The CalypsoAI Inference Platform tests, defends, and monitors AI in development and production. With Defend, Red-Team, and Observe,
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM10
- Open Source
- Stage: Deploy, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Noma Security
- Adversarial Attack Protection, Adversarial Input Detection, Adversarial Testing, AI/ML Bill of Materials (BOM), Automated Vulnerability Scanning, Code Analysis, Data Privacy and Protection, GenAI/LLM Posture Management, Incident Simulation, LLM & App Vulnerability Scanning, LLM Enabled Web Application Firewall, LLM Guardrails, Model and Application Interaction Security, Penetration Testing, Prompt Security, Response Testing, Runtime Application Self-Protection, SAST/DAST/IAST, Secure CI/CD pipeline, Secure Coding Practices, Secure Library/Code Repository, Secure Output Handling, Software Composition Analysis, Supply Chain Verification, Third-Party Risk Assessment, Vulnerability Scanning
Noma Security
"Noma Security is a comprehensive application security solution for the Data and AI lifecycle. It offers , End-to-End Visibility: Scanning notebooks, source code, and other
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Govern
Cranium Platform and AI Trust Hub
Cranium
Whether organizations are builders and/or consumers of AI, Cranium offers a comprehensive platform that enables complete security, compliance, and trust across the entire AI supply
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Govern, Monitor
Infotect Security
IWS scans outbound response traffic in real time for undesirable content and confidential data at layer 4. It is a paradigm shift in web security,
- LLM01, LLM02, LLM05, LLM06, LLM07
- Proprietary
- Stage: Govern, Monitor, Operate, Test & Evaluate
DynamoGuard
- Adversarial Attack Protection, Adversarial Testing, Compliance Management, Data Leakage Protection, Ethical Compliance, LLM Benchmarking, LLM Guardrails, Observability, Penetration Testing, Prompt Security, Regulatory Compliance Tracking, Risk Assessment and Management, User Activity Monitoring, Vulnerability Scanning
Dynamo AI
DynamoGuard offers real-time guardrailing for GenAI, customizable in natural language and capable of running in the cloud, hybrid, on-prem or fully on edge devices to
- LLM01, LLM06, LLM09
- Proprietary
- Stage: Govern
Unbound Security
Unbound Security
Unbound AI gateways solves for guardrails, prompt injection, and jailbreaking attacks while helping customers create routing policies based on data sensitivity. For example, prompts containing
- LLM01, LLM02, LLM05, LLM08
- Proprietary
- Stage: Deploy, Govern, Monitor, Operate, Release
Palo Alto Networks AI Runtime Security
- Adversarial Input Detection, AI/ML Bill of Materials (BOM), Compliance Management, Compliance Verification, Data Privacy and Protection, Incident Governance, Network Security Validation, Observability, Regulatory Compliance Tracking, Risk Assessment and Management, Security Alerting, Security Metrics Collection, User and Data Privacy Protections
Palo Alto Networks
Palo Alto Networks AI Runtime Security provides continuous discovery, protection, and monitoring for genAI applications, preventing security risks such as prompt injections, sensitive data leakage,
- LLM01, LLM02, LLM03, LLM04, LLM06, LLM07, LLM08, LLM09, LLM10
- Proprietary
- Stage: Govern, Monitor, Operate
Blueteam AI Gateway
- Adversarial Attack Protection, Bias and Fairness Oversight, Compliance Management, Data Integrity and Encryption, Data Leakage Protection, Data Security Posture Management, LLM Guardrails, Privacy, Prompt Security, Runtime Application Self-Protection, Secure Output Handling, User/Machine Access audits
Blueteam AI
Blueteam AI Gateway is a network-layer appliance that intercepts traffic to AI models and discovers AI use, safeguards data from leaking, and governs safe and
- LLM01, LLM04, LLM06, LLM09
- Proprietary
- Stage: Govern, Monitor, Operate
Aim AI Security Platform
- Adversarial Attack Protection, Automated Vulnerability Scanning, Compliance Management, Data Leakage Protection, Data Security Posture Management, LLM Guardrails, LLM Incident Detection and Response, Privacy, Prompt Security, Risk Assessment and Management, Runtime Application Self-Protection, Secure Output Handling, User/Machine Access audits
Aim Security
The Aim AI Security Platform enables enterprises to secure every AI interaction throughout their AI adoption journey, from AI applications used directly by employees to
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08
- Proprietary
- Stage: Develop & Experiment, Govern, Monitor, Release, Test & Evaluate
Cisco Systems
Cisco AI Validation assesses AI applications and models for security and safety vulnerabilities. We automatically analyze a model’s risk across hundreds of attack techniques and
- LLM01, LLM03, LLM04, LLM06, LLM09
- Proprietary
- Stage: Data Augmentation and Fine-Tuning, Develop & Experiment, Govern, Monitor, Operate, Release, Scope & Plan, Test & Evaluate
Pillar Security
- Adversarial Attack Protection, Adversarial Testing, AI/ML Bill of Materials (BOM), Data Leakage Protection, Early Identification of Sensitive Data, Ethical Compliance, Incident Simulation, LLM & App Vulnerability Scanning, LLM Benchmarking, LLM Guardrails, Model and Application Interaction Security, Model Behavior Analysis, Penetration Testing, Privacy, Regulatory Compliance Tracking, Response Testing, Runtime Application Self-Protection, Security Alerting, Software Composition Analysis, User Activity Monitoring
Pillar Security
Pillar enables teams to rapidly adopt AI with minimal risk by providing a unified AI security layer across the organization
- LLM01, LLM02, LLM04, LLM06, LLM07, LLM08
- Open Source
- Stage: Govern
AI Verify Foundation
AI Verify is an AI governance testing framework and software toolkit that validates the performance of AI systems against a set of internationally recognized principles
- Proprietary
- Stage: Augment/Fine Tune Data, Deploy, Develop & Experiment, Govern, Monitor, Operate, Scope & Plan
Securiti
Securiti Data Command Center provides unified intelligence, controls, and orchestration for enabling the safe use of data and AI across hybrid multi-clouds. Enterprises rely on
- LLM01, LLM02, LLM03, LLM04, LLM05, LLM06, LLM07, LLM08, LLM09, LLM10
