Solution approach
AI testing and adversarial assurance
Test models, AI applications, knowledge-retrieval systems, and agents for failures and adversarial behavior before and after release.
View research coverageThese counts describe available public research, not product quality or suitability.Expand
- Vendor profiles
- 16 shown in this view
- Security requirements
- 3 to review
- Strong public support
- 31 requirement records
- Limited public support
- 11 requirement records
- Open research
- 6 with no supporting claim · 0 incomplete
Evaluation guide
What buyers should verify
AI assurance and adversarial testing
Test models, applications, retrieval-augmented generation (RAG) systems, and agents before release and continuously thereafter using adversarial probes, evaluation suites, attack simulation, and security release gates.
Questions to test
- A controlled test campaign exercises an AI model, application, or agent against named AI attack classes.
- Results include reproducible prompts or attack steps, affected component, severity, and remediation guidance.
- A retest or release gate records whether the identified weakness was remediated.
AI model and supply-chain security
Discover, inventory, scan, validate, and monitor models, datasets, model artifacts, registries, dependencies, and AI development assets for tampering, unsafe serialization, provenance gaps, or malicious content.
Questions to test
- A test model or AI artifact appears in inventory with origin, version, hash or provenance, and deployment context.
- A malicious, tampered, unsafe, or policy-violating artifact produces a finding before deployment.
- The finding links the artifact to its registry, pipeline, owner, affected workload, and remediation action.
Generative AI application security
Protect enterprise-built large language model (LLM) applications, retrieval-augmented generation (RAG) systems, prompts, application programming interfaces (APIs), model calls, tools, and production runtime behavior.
Questions to test
- A test large language model (LLM) application event records prompt, application programming interface (API), model, retrieval, or tool interaction context.
- A prompt-injection or unsafe-output test is detected, blocked, or flagged by the guardrail or large language model (LLM) firewall.
- The evidence links the event to an application, endpoint, request identifier, or workflow.
Vendor research
Vendors with public research for this approach
Vendors are grouped by whether this approach is a core product focus or related coverage, then ordered by documented support across the requirements shown. This organizes public research coverage; it is not a product ranking or recommendation.
Evidence labelsStrong public supportLimited public supportNo supporting claim foundResearch incomplete
VendorAI assuranceModel supply chainGenerative AI app securitySourcesAction
Core product focusThis approach is central to how these vendors present the product · 12 vendors
Cisco AI DefenseEstablishedSupport found for 3 of 3 requirementsStrong public supportStrong public supportStrong public support2Profile →
HiddenLayerGrowth stageSupport found for 3 of 3 requirementsStrong public supportStrong public supportStrong public support3Profile →
Palo Alto Networks Prisma AIRSEstablishedSupport found for 3 of 3 requirementsStrong public supportStrong public supportStrong public support3Profile →
AktoEmergingSupport found for 3 of 3 requirementsStrong public supportLimited public supportStrong public support3Profile →
Enkrypt AIEmergingSupport found for 3 of 3 requirementsStrong public supportLimited public supportStrong public support3Profile →
Holistic AIEmergingSupport found for 3 of 3 requirementsStrong public supportLimited public supportStrong public support1Profile →
MindgardEmergingSupport found for 3 of 3 requirementsStrong public supportLimited public supportStrong public support3Profile →
Snyk EvoScaledSupport found for 3 of 3 requirementsStrong public supportStrong public supportLimited public support1Profile →
F5 AI Security PlatformEstablishedSupport found for 2 of 3 requirementsStrong public supportNo supporting claim foundStrong public support3Profile →
Harness AI Security / TraceableEstablishedSupport found for 2 of 3 requirementsStrong public supportNo supporting claim foundStrong public support1Profile →
LatticeFlow AIEmergingSupport found for 2 of 3 requirementsStrong public supportNo supporting claim foundStrong public support1Profile →
Lakera / Check PointEstablishedSupport found for 2 of 3 requirementsNo supporting claim foundLimited public supportStrong public support3Profile →
Related coverageThese vendors address the requirements through another core product focus · 4 vendors
Upwind AI SecurityScaledSupport found for 3 of 3 requirementsStrong public supportLimited public supportStrong public support1Profile →
Zscaler AI SecurityEstablishedSupport found for 3 of 3 requirementsStrong public supportLimited public supportStrong public support2Profile →
Credo AIGrowth stageSupport found for 3 of 3 requirementsStrong public supportLimited public supportLimited public support2Profile →
ModelOpEmergingSupport found for 1 of 3 requirementsLimited public supportNo supporting claim foundNo supporting claim found1Profile →
Company maturity remains a filter and profile attribute; it does not affect the research-coverage order.