“One of the biggest security reforms we are announced is within us AI protection Solution (part of a customer’s GCP Safety Command Center). As organizations fast AI, we are developing new abilities to help protect our initiatives. ,
These include new abilities for automatic discovery of AI agents and Model reference protocol (MCP) Server at Security Command Center. Makhani said that this update, which will soon be available in preview, will help the defenders automatically identify their AI agent ecosystem automatically in the ecosystem; Prevention of new agent-specific runtime risks such as tool poisoning and indirect early injections; And increase the reaction of the event for risk -filled behavior and external hazards such as discrepancy on agent assets and suspected behavior.
Makhani also said that the new capabilities declared will help in promoting the vision of an agent security operations center (Agent soak) That customers can manufacture through GCP. An agent SOC AI allows agents to coordinate their tasks to meet a shared target.
