Salt Security Enhances Partnership with CrowdStrike to Improve AI Agent Security
In a significant advancement for cybersecurity, Salt Security has announced an expansion of its integration with CrowdStrike, aimed at enhancing the visibility that security teams have over the connections between AI agents and enterprise systems. This collaboration merges Salt’s Agentic API platform with CrowdStrike’s comprehensive suite, including the Falcon Foundry, Falcon Next-Gen SIEM, and Falcon Firewall Management. The integration is designed to assist joint customers in identifying AI agents operating within their environments, enabling them to monitor the infrastructure and permissions that empower these agents to perform their functions.
This strategic move comes at a pivotal time as many enterprises transition from merely experimenting with generative AI to deploying autonomous agents capable of executing tasks independently. These agents possess the ability to hold credentials, connect to internal systems through Model Context Protocol (MCP) servers and a variety of tools, invoke APIs, and execute actions on behalf of employees, all of which poses new challenges for maintaining security.
The challenges faced by security teams have grown more complex as AI agents increasingly interact with tools, APIs, and MCP servers that may not have undergone proper security review processes or are afforded broader permissions than necessary. Salt’s expansion allows customers to utilize a certified application on the CrowdStrike Falcon Foundry, which facilitates the deployment of Salt’s solutions using existing Falcon sensors, thereby circumventing the need for additional gateways or proxies that could complicate the security architecture.
Through this enhanced integration, organizations can effectively identify AI agents and discern the MCP servers and tools they connect to, as well as the APIs invoked by these connections. Furthermore, it assists in identifying publicly exposed MCP servers and integrations that may have been introduced without following the established security protocol of the organization.
Salt Security’s findings can also be aggregated within CrowdStrike’s Falcon Next-Gen SIEM, allowing for the correlation of data with existing telemetry related to endpoints, identity, and cloud activities. This robust synthesis of information assists organizations in detecting deviations in agent behavior, enabling them to employ Falcon Firewall Management as a mechanism for automated response actions.
The companies underscore the importance of addressing three pivotal questions for security teams: which AI agents are currently operational within the organization, what systems and data are accessible to these agents, and whether their actual behavior aligns with the permissions granted. Roey Eliyahu, co-founder and CEO of Salt Security, emphasized that while most enterprises can identify approved AI agents, far fewer are capable of tracking the complete path these agents take once they are active within the business environment. He noted that while agents may initiate their activities based on legitimate prompts, risks can develop over time through over-permissioned tools, MCP connections, or API calls. By uniting Salt’s API capabilities with the AI-native features of the Falcon platform, they aim to provide customers with greater visibility and oversight of agent operations across their enterprises.
The announcement signifies an expansion of the already established partnership between the two cybersecurity companies. In 2022, CrowdStrike’s Falcon Fund invested in Salt Security, which led to the rollout of a certified Salt application on the Falcon Foundry and subsequent integration with Falcon Next-Gen SIEM.
This latest development not only fortifies their existing relationship but also extends their collaborative efforts specifically into the realm of AI agent security. With organizations granting autonomous systems greater access to vital business applications, data repositories, and operational workflows, the need for heightened security measures has never been more critical.
The collaboration between Salt Security and CrowdStrike stands as a testament to the evolving landscape of cybersecurity, particularly as it pertains to AI technology. As AI continues to permeate various facets of business operations, safeguarding these digital agents becomes imperative for maintaining overall organizational security. This integration offers a blueprint for enterprises looking to harness the power of AI while mitigating the associated risks effectively.
In conclusion, as AI’s role within corporate environments grows, the partnership between Salt Security and CrowdStrike will likely play an instrumental role in shaping the future of cybersecurity. Security teams will find themselves better equipped to monitor, manage, and respond to the actions of AI agents, thereby fostering safer operational frameworks across the board.