Check Point integrates AI Agent Security with Google Cloud Gemini Enterprise Agent Registry and Agent Gateway
Check Point announced that its AI Agent Security capabilities integrate with Google Cloud's Gemini Enterprise, adding AI-specific risk assessment and inline runtime protection to the platform's centralized agent control plane. The integration targets the risk that an authenticated and authorized agent can still be manipulated into harmful actions, for example through prompt injection. Check Point is a launch security partner for Gemini Enterprise, announced at Google Cloud Next '26, and the capability is powered by the Check Point AI Defense Plane. This is vendor product news with no specific vulnerability, incident or threat actor described.
Technical Details
- Agent Registry: Check Point observability reads the registry to surface risk posture on agents and their connected tools, including risky configurations, permissions and Model Context Protocol (MCP) connections, before workflows start.
- Agent Gateway prompt defense: user-to-agent and Agent2Agent (A2A) traffic is vetted inline by prompt injection defenses embedded as Agent Gateway service extensions, before the agent acts.
- MCP security: MCP tool calls are catalogued and permissions are checked against agent identity. Indirect prompt injection carried in MCP server content is blocked in real time.
- Data protection: the gateway inspects interactions to identify sensitive data and enforces policy before information is returned, transferred or passed to a downstream tool.
- Context-aware evaluation: prompts, responses, tool calls and multi-step behavior are assessed, not just whether the agent holds permission for an action.
- Vendor-reported metrics: at Dropbox, 99.8% of real-world threats blocked, sub-40ms average latency and a false positive rate under 0.5%. At Nubank, a 0.01% false positive rate across 100+ languages. These figures are vendor claims and are not independently verified.
Remediation
The article presents no vulnerability to patch. Its guidance is to inventory the agent estate, assess agent and MCP connection risk before production, set policy boundaries, and inspect agent behavior inline at runtime.
Linha do tempo
-
Check Point integrates AI Agent Security with Google Cloud Gemini Enterprise Agent Registry and Agent Gateway
ler a análise desta fonte ocultar
Check Point announced that its AI Agent Security capabilities integrate with Google Cloud's Gemini Enterprise, adding AI-specific risk assessment and inline runtime protection to the platform's centralized agent control plane. The integration targets the risk that an authenticated and authorized agent can still be manipulated into harmful actions, for example through prompt injection. Check Point is a launch security partner for Gemini Enterprise, announced at Google Cloud Next '26, and the capability is powered by the Check Point AI Defense Plane. This is vendor product news with no specific vulnerability, incident or threat actor described.
Technical Details
- Agent Registry: Check Point observability reads the registry to surface risk posture on agents and their connected tools, including risky configurations, permissions and Model Context Protocol (MCP) connections, before workflows start.
- Agent Gateway prompt defense: user-to-agent and Agent2Agent (A2A) traffic is vetted inline by prompt injection defenses embedded as Agent Gateway service extensions, before the agent acts.
- MCP security: MCP tool calls are catalogued and permissions are checked against agent identity. Indirect prompt injection carried in MCP server content is blocked in real time.
- Data protection: the gateway inspects interactions to identify sensitive data and enforces policy before information is returned, transferred or passed to a downstream tool.
- Context-aware evaluation: prompts, responses, tool calls and multi-step behavior are assessed, not just whether the agent holds permission for an action.
- Vendor-reported metrics: at Dropbox, 99.8% of real-world threats blocked, sub-40ms average latency and a false positive rate under 0.5%. At Nubank, a 0.01% false positive rate across 100+ languages. These figures are vendor claims and are not independently verified.
Remediation
The article presents no vulnerability to patch. Its guidance is to inventory the agent estate, assess agent and MCP connection risk before production, set policy boundaries, and inspect agent behavior inline at runtime.