Public-source research
Vendor evidence
Review what vendors say publicly, the exact quoted source text, the security requirement each statement may support, and what still needs verification.
Research library coverageCounts describe the research workflow, not vendor quality or product effectiveness.View details
Evidence records1280
Source checked863
Needs verification0
Source captured0
No supporting claim found411
Excluded from evidence6
Operant AI claims inline controls that block rogue agents, out-of-scope actions, unauthorized access, and critical agentic attacks.
blocking out-of-scope actions before execution and alerting your security team with full action traces before damage is done.
Related framework references (5)
Tenable One AI Exposure claims prioritized actions to fix AI misconfigurations, enforce AI acceptable-use policies, and reduce AI exposure.
Take action with prioritized insights to fix AI misconfigurations, enforce AI acceptable use policies (AI AUP), and reduce AI exposure.
Related framework references (5)
No public claim found for this capability.
No quoted source text is recorded for this claim.
Related framework references (5)
Noma claims runtime policy enforcement for AI and agentic communication.
See all AI and agentic communication as it happens and enforce policies at runtime.
Related framework references (5)
Onyx claims fine-grained language-model provider access controls and per-team model approvals inside Onyx, which partially maps to unapproved AI control.
Onyx provides fine-grained access control for language model providers, allowing administrators to control **who** can use specific models and **which agents** can use them.
Related framework references (5)
CrowdStrike claims it can block unauthorized AI access and topics while masking or encrypting sensitive data before exposure.
Block unauthorized access and topics or mask or encrypt sensitive data before exposure
Related framework references (5)
HiddenLayer claims AI Guardrails enforce policies that prevent prompt injection, data leakage, and unsafe AI behavior in real time.
AI Guardrails Enforce policies that prevent prompt injection, data leakage, and unsafe AI behavior in real time.
Related framework references (5)
Aurascape claims it applies policy based on role, sensitivity, and conversation context.
Apply policy based on role, sensitivity, and conversation context
Related framework references (5)
Cyberhaven claims runtime guardrails block high-risk data movement, redirect users to approved tools, and coach employees.
Enforces runtime guardrails at the prompt and response level, blocking high-risk data movement, redirecting users to sanctioned tools, and coaching employees with plain-English policy explanations.
Related framework references (5)
Knostic claims it detects shadow AI, blocks data exfiltration, and stops destructive commands.
We detect shadow AI, block data exfiltration, and stop destructive commands like rm -rf .
Related framework references (5)
Island claims data boundaries and data loss prevention (DLP) can prevent corporate data from being entered into unapproved AI apps and services.
Using data boundaries and data loss prevention, Island can ensure your corporate data doesn’t get inputted into unsanctioned AI apps and services.
Related framework references (5)
Aembit claims the ability to stop AI access and audit access in real time based on an agent's unique identity.
Stop AI access with a click of a button. Audit access in real-time — based on the agent’s unique identity, even if it’s operating on behalf of a user.
Related framework references (5)
Wiz materials reviewed did not provide a public claim for employee unapproved AI app blocking, coaching, allowlists, browser control, or network control.
No quoted source text is recorded for this claim.
Related framework references (5)
Varonis claims controls to revoke excessive permissions, fix risky AI misconfigurations, and keep sensitive data out of large language models (LLMs).
Visualize AI’s sensitive data access Revoke excessive permissions Fix risky AI misconfigurations Monitor AI-created data Classify AI-generated content Apply sensitivity labels Monitor prompts and alert on suspicious activity Keep sensitive data out of LLMs Discover hidden AI workloads Identify sensitive data flows Map AI accounts with access to sensitive data
Related framework references (5)
Pangea claims Prompt Guard analyzes user and system prompts to block jailbreak attempts and organizational limit violations.
Pangea Prompt Guard analyzes user and system prompts to block jailbreak attempts and organizational limit violations.
Related framework references (5)
F5 claims AI Security Platform translates risk, privacy, and compliance obligations into enforceable policy for enterprise AI use.
Translate risk appetite, privacy requirements, and compliance obligations into enforceable policy for enterprise AI use.
Related framework references (5)
Iterate.ai claims AgentWatch can enforce AI governance policy outcomes with block, warn, or allow actions and evidence trails.
Block, warn, or allow with full evidence trails
Related framework references (5)
Credo AI claims policy enforcement and production monitoring, with planned enforcement integrations for CI/CD, cloud access security broker (CASB), and application programming interface (API) gateways.
Planned enforcement integration with CI/CD, CASBs, and API gateways
Related framework references (5)
Holistic AI claims automated blocking, kill switches, access revocation, and human-review escalation when policy violations or anomalies are detected.
blocking unsafe outputs, activating kill switches, revoking access or flagging for human review
Related framework references (5)
Mindgard materials reviewed did not provide a public claim for workforce allow, coach, restrict, or block policy over unapproved AI use.
No quoted source text is recorded for this claim.
Related framework references (5)