Exclusive: FireMon unveils AI agents to handle the legwork behind firewall policy changes

FireMon, a network security policy management company, has introduced AI agents for its platform to assist with the majority of tasks related to firewall policy changes. The new feature, known as FireMon Agentic AI, aims to manage the expanding workload that comes with each firewall vendor and cloud environment an enterprise operates.

A single business access request might involve multiple devices and policies, with several technically valid ways to grant access. Traditionally, engineers have handled these decisions manually, taking up valuable expertise as businesses seek faster decisions.

With FireMon Agentic AI, agents manage the investigation and planning stages, operating within limits set by the customer regarding which devices and tools each one can access. The data used for this work is sourced from the same policy records that human users query. While AI-generated choices requiring organizational judgment are routed back to practitioners, the FireMon Intelligent Agent can answer plain-language questions about policy, such as which rules permit access to a specific service, generating the underlying queries itself.

The Network Designer Agent provides firewall rule recommendations for a given access requirement, and the Policy Change Designer Agent carries an access request from discovery to staged Policy Plan changes, automatically filling in required fields like rule names and pausing when a human must choose between valid alternatives. The proposal only gets stored in FireMon's workflow after review and acceptance, with no automatic deployment to firewalls.

Additionally, authorized external workflows can utilize the FireMon Model Context Protocol server to pull policy evidence alongside alert data.

FireMon CEO Jody Brazil stated that network security has become increasingly complex, and the solution is not more infrastructure but rather putting the accumulated policy context to work while reserving human judgment for critical decisions. AI enablement is optional, defaulting to off, with customers able to choose from supported large language models, including one hosted in their approved environment.

Each agent's permissions can be scoped to particular device groups, and activity is logged for an audit trail. FireMon Agentic AI and the FireMon MCP Server are set to launch in November, and FireMon does not charge extra for the initial release. The company, founded in 2004 by Jody Brazil, currently has over 1,800 enterprise customers in more than 70 countries.