What is Check Point AI Security?

Security Framework Coverage

Frameworks such as the OWASP Top 10 for LLMs, the OWASP Top 10 for agentic applications, and MITRE ATLAS describe categories of risk. Most categories are addressed by more than one capability, across the posture and runtime layers, and some depend on controls in your own architecture. The tables below show, for each framework item, which layer of AI Agent Security addresses it and what remains the responsibility of your architecture.

How to read the mapping: for each item, the first column shows the runtime contribution (the Guard API guardrails), the second the posture contribution (discovery and risk assessment), and the third the controls that sit with your own architecture. An empty cell means that item is not the primary place that layer contributes.

OWASP Top 10 for LLMs (2025)

ItemRuntimePostureYour architecture
LLM01 Prompt InjectionPrompt Defense screens user input, tool responses, tool descriptions, and untrusted contentFlags configuration that widens the injection surface, such as untrusted content sourcesSeparate trusted instructions from untrusted data; scope user permissions
LLM02 Sensitive Information DisclosureData Leakage detection on outputs; Prompt Defense on extraction attemptsFlags agents that combine sensitive-data access with exposure pathsData minimization; scope what data tools can reach
LLM03 Supply ChainFlags unofficial, unknown, and vulnerable MCP servers and suspicious or likely-malicious tool codeVet and pin components
LLM04 Data and Model PoisoningScreens retrieved content and tool responses; screen content at ingestionValidate training and grounding data; ingestion controls
LLM05 Improper Output HandlingDetects suspicious outputs such as unknown linksSanitize and encode outputs before downstream use
LLM06 Excessive AgencyAgent Behavior Defense: Off-Task Action detection and Tool Allow/Deny ListFlags excessive tools, unused tools, and high-risk tool combinationsLeast-privilege tool scoping; human approval gates
LLM07 System Prompt LeakageData Leakage detection on outputs; Prompt Defense on extraction attemptsKeep secrets out of system prompts
LLM08 Vector and Embedding WeaknessesScreens retrieved RAG content and tool responsesRAG access controls and tenant isolation
LLM09 MisinformationContent Moderation and custom guardrailsGrounding and verification in the application
LLM10 Unbounded ConsumptionCustom guardrails detect some patternsRate limits and quotas

OWASP Top 10 for Agentic Applications (2026)

ItemRuntimePostureYour architecture
ASI01 Agent Goal HijackPrompt Defense across inputs, tool responses, and tool descriptions; Off-Task Action detectionFlags untrusted content sources and missing isolationSeparate instructions from data; least privilege
ASI02 Tool Misuse and ExploitationOff-Task Action detection; Tool Allow/Deny ListFlags write-access tools, high-risk connections, and risky tool combinationsStrict tool scoping and input validation
ASI03 Identity and Privilege AbuseTool Allow/Deny List limits available actionsFlags missing authentication, static credentials, and author-credential execution where the platform exposes itIdentity and access management; least privilege per agent
ASI04 Agentic Supply ChainFlags unofficial, unknown, and vulnerable MCP servers, untrusted components, and suspicious or likely-malicious tool codeComponent vetting and provenance
ASI05 Unexpected Code ExecutionOff-Task Action detection and tool controlsSandboxing; disable code execution where it is not needed
ASI06 Memory and Context PoisoningScreens retrieved content and tool responses; screen content at ingestionMemory and session integrity; validation
ASI07 Insecure Inter-Agent CommunicationScreens content passed between agents when integrated into the flowAuthenticated and validated inter-agent channels
ASI08 Cascading FailuresOff-Task Action detection and tool controls limit propagationIsolation and circuit breakers
ASI09 Human-Agent Trust ExploitationContent Moderation; Prompt DefenseHuman oversight in the workflow
ASI10 Rogue AgentsOff-Task Action detection and tool controlsDiscovery and risk assessment surface unknown, unmanaged, and over-privileged agentsGovernance and decommissioning

Notes on coverage

  1. Inter-agent communication (ASI07) is addressed indirectly: by inspecting the content agents pass between each other when those interactions are screened through the Guard API. Cryptographic and identity controls for inter-agent channels sit with your architecture.
  2. Runtime protection today integrates through the Guard API. Native runtime integration with agent platforms is on the roadmap.
  3. Posture-level enforcement of agent component allow/deny lists is on the roadmap. Runtime tool access control is available today through the Tool Allow/Deny List in Agent Behavior Defense.
  4. MITRE ATLAS techniques are mapped on individual risks in the risk assessment, alongside the OWASP references shown in the risk-types view.