Aviatrix launches containment platform for AI workloads

Aviatrix announced a new Containment Platform for agentic AI on April 29, 2026, extending its Cloud Native Security Fabric with two components: Zero Trust for AI Workloads, now generally available, and Aviatrix AgentGuard, now in early access, per a GlobeNewswire press release. The company described the platform as enforcing "Communication Governance" across VMs, Kubernetes clusters, and serverless functions without agents or code changes, according to the release. Reporting and interviews with Aviatrix executives in SiliconANGLE and CRN emphasize a shift from detection to limiting blast radius; Aviatrix CEO Doug Merritt told SiliconANGLE, "My argument for the containment era is the most important metric is blast radius." CRN reported that AgentGuard discovers authorized and shadow agents, maps connected LLMs, tools, and data, and builds continuous risk profiles.
What happened
Aviatrix launched a Containment Platform for agentic AI workloads on April 29, 2026. Per a GlobeNewswire press release from Aviatrix, the launch bundles Zero Trust for AI Workloads, which the company says is now generally available, and Aviatrix AgentGuard, which the company says is now in early access. The GlobeNewswire release describes the offering as the "industry's first" Containment Platform for AI agents. The release states the platform enforces "Communication Governance" across VMs, Kubernetes clusters, and serverless functions without deploying agents or requiring code changes. The press release cites industry research claiming only 5 to 20 percent of enterprises have controls in place to limit where a workload can communicate, a gap the company frames as central to agentic risk. The release also references a 2026 supply chain campaign called the Cascade, attributed in the release to TeamPCP, which it says affected 36 percent of enterprise cloud environments at the time of compromise.
Technical details
CRN reports that Aviatrix AgentGuard discovers every agent running across virtual machines, Kubernetes clusters, and serverless functions, including both authorized and "shadow" agents, and maps the large language models, tools, and data the agents connect to while building a continuous risk profile. The GlobeNewswire release and CRN say the platform extends the Aviatrix Cloud Native Security Fabric to enforce per-workload communications policies and block common exfiltration vectors by default. SiliconANGLE and CRN carried quotes from Aviatrix executives that emphasize containment and blast radius as the operational metric motivating the product. Doug Merritt, Aviatrix CEO, told SiliconANGLE, "My argument for the containment era is the most important metric is blast radius." Chris McHenry, Chief Product Officer, is quoted in the GlobeNewswire release describing a single enforcement principle: each workload can only communicate with what it is explicitly permitted to reach.
Editorial analysis
Industry-pattern observations: Enterprises are moving from perimeter- and detection-focused controls toward workload-level constraints as agentic systems proliferate. Containment architectures shift security tradeoffs by treating lateral movement and interservice reach as primary controls rather than downstream alerts. For practitioners, this raises integration questions: enforcement at the workload layer must interoperate with cloud provider networking primitives, service meshes, identity systems, and observability pipelines. Channel coverage and managed services will matter for organizations with limited internal integration capacity, a point CRN highlights when quoting Aviatrix on partner opportunities.
Context and significance
Editorial analysis: The product launch reflects a broader market response to two converging trends: wider deployment of agentic AI inside enterprise workloads and an expanding threat surface where compromised agents can cause cross-service impact. Vendors and integrators have increasingly framed "blast radius" as an operational security metric, and Aviatrix is packaging that framing into a network- and workload-level product. This matters to security and cloud engineering teams because containment affects how teams design service-to-service policies, secrets management scope, and incident response playbooks. The platform claim of operating without agents or code changes targets a common adoption barrier, but real-world effectiveness will depend on policy coverage, false-positive rates, and integration friction with existing CI/CD and observability tooling.
What to watch
For practitioners: observers should watch:
- •how broadly Aviatrix opens AgentGuard beyond early access and which cloud provider APIs and service meshes it supports
- •independent validation of discovery and mapping accuracy for both authorized and shadow agents
- •how the product integrates policy decisions with existing identity and secrets stores. Observers should also watch partner announcements and managed-service offerings that arise, as CRN flagged channel revenue opportunities tied to containment assessments and deployment services
Direct quote context
CRN and SiliconANGLE carried quotes from Aviatrix executives emphasizing blast radius as the key metric for containment. Per the GlobeNewswire release, Aviatrix framed the Cascade campaign and the lack of workload-level enforcement as motivators for the new platform.
Scoring Rationale
This is a notable product launch addressing a tangible operational risk for cloud and security teams: agentic AI lateral movement and exfiltration. It is not a paradigm-shifting model or regulation, but it introduces a practical containment approach that practitioners will evaluate for integration and efficacy.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
