Gen (NASDAQ: GEN), the parent company of Norton, Avast, LifeLock, and MoneyLion, announced on April 30 two products that extend consumer security into the AI agent era: a VPN designed from scratch for autonomous agents and an expanded Norton AI Agent Protection suite integrated into Norton 360.
The VPN for Agents is, according to Gen, the first consumer-grade VPN built specifically for AI agents rather than human web browsing. Norton AI Agent Protection, meanwhile, now monitors supported agents including Claude Code, Cursor, and OpenClaw, adding prompt injection defense and malicious tool scanning to the Norton 360 suite used by tens of millions of customers.
How the Agent VPN Works
Traditional VPNs route all device traffic through a single encrypted tunnel. They cannot distinguish between a human browsing the web and an AI agent making API calls to sensitive enterprise systems. Gen’s VPN for Agents addresses this with three design choices, according to the press release:
No client installation required. The VPN operates without downloads or local setup, reflecting how agents run as cloud processes rather than local applications.
Multi-tunnel architecture. Agents can operate simultaneously across different countries through separate tunnels. This is a first for consumer VPN products and addresses the reality that autonomous agents often need to access region-specific services concurrently.
Traffic isolation. Agent traffic is separated from human traffic, giving users visibility into what their agents are doing and where they are connecting, independent of their own browsing.
“As people embrace AI agents and use them to manage more of their digital lives, they deserve security and privacy that keeps pace,” Howie Xu, Gen’s Chief AI Innovation Officer, said in the announcement.
Norton 360 Agent Protection Expansion
The Norton AI Agent Protection expansion brings action-level monitoring to the agents consumers are actually using. The system watches what supported agents do and where they connect, inserting security layers between an agent’s decision and its execution.
New capabilities include pre-use scanning of AI plugins, skills, and tools to block malicious or over-privileged integrations before an agent can use them. Prompt injection defense detects and neutralizes attacks that try to manipulate agent behavior. Advanced code and file scanning analyzes content that agents access or generate, catching malware and unsafe scripts before execution.
Norton AI Agent Protection is currently integrated into Norton 360 for Windows, with Mac support coming, according to Gen. Both products are available through Gen’s Agent Trust Hub, which now spans verification, detection, and real-time communication security.
Consumer vs. Enterprise Agent Security
Gen’s positioning is deliberate. While Palo Alto Networks, CrowdStrike, and Aviatrix are building enterprise agent governance platforms, Gen is targeting the individual user who has Claude Code writing their application or OpenClaw managing their workflows. The consumer angle matters because agent adoption is not waiting for enterprise procurement cycles. Individual developers and knowledge workers are deploying agents now, often without any security layer between the agent and their sensitive data.
The question for Gen is scale. Millions of Norton 360 subscribers represent immediate distribution, but consumer agent security is an unproven category. Whether users will pay for agent-specific protection, or even understand why they need it, remains to be tested.