650 Group Research: Hybrid AI Drives Growth in the Deep Observability Market

AI is changing more than applications. It is changing the infrastructure underneath them.

AI agents, machine-to-machine interactions, data-intensive workloads, and new AI services are generating more data in motion across enterprise networks. At the same time, organizations are adopting a growing mix of AI models and services across public cloud, private infrastructure, SaaS, and specialized AI environments.

This shift to Hybrid AI changes what organizations need from observability. As networks get faster, infrastructure becomes more distributed, and the attack surface expands, simply collecting more telemetry isn’t enough. Organizations need to extract the right signal and context from rapidly growing volumes of data and efficiently deliver that intelligence to the security, observability, data, and AI platforms that depend on it.

Our new research at 650 Group provides evidence of that shift. According to our latest report, the deep observability market grew 22 percent year over year in the first half of 2026, and we project a 28 percent CAGR through 2030, reaching $2 billion. Gigamon maintained its leadership position with 51 percent market share, while its deep observability revenue grew 24 percent year over year.

We believe acceleration of the deep observability market is the result of four key drivers:

1.  AI Is Accelerating Network Scale

AI isn’t just another application workload. Applications, agents, models, APIs, and automated workflows are creating a growing volume of machine-generated traffic as machines increasingly communicate with applications, data, models, tools, and other agents.

That growth is adding momentum to the transition toward 400G and 800G infrastructure. But as network capacity increases, organizations cannot allow the cost and complexity of observing that infrastructure to increase at the same rate. Doubling network capacity cannot mean simply doubling telemetry volumes, infrastructure footprint, storage, and downstream processing costs.

Observability architectures instead need to become more efficient at extracting what matters from network traffic. Network traffic provides a rich, independent source of evidence, while application metadata can distill massive traffic volumes into high-value context that security and IT teams can use at scale.

2. Hybrid AI Fragments Infrastructure Visibility

The infrastructure supporting AI is also becoming more diverse. Organizations are using different AI models and environments based on economics, performance, security, governance, sovereignty, and workload requirements. Frontier models may run in public cloud services, while open-weight or specialized models operate in private infrastructure or other environments.

This emerging Hybrid AI architecture creates more connections among models, agents, applications, enterprise data, and infrastructure, which in turn creates more activity for security and IT teams to understand.

Visibility cannot stop at the boundary between public and private cloud, or between traditional applications and AI workloads. Organizations need a consistent view of data in motion as workloads, applications, agents, and data traverse increasingly diverse environments and protocols.

AI may be hybrid, but visibility can’t be fragmented.

3. AI Expands the Security Surface

Every new AI application, model, agent, API, identity, and communication path creates activity security teams need to discover, understand, and govern. Shadow AI adds another layer of complexity, while agentic AI means security teams increasingly need to understand not only who a user is, but what agents are doing on behalf of users and applications.

Emerging protocols such as Model Context Protocol (MCP) add new communication paths connecting AI applications and agents with models, tools, and enterprise systems.

The network provides an independent view of this activity and data in motion. Network-derived telemetry can help organizations identify AI activity, understand lateral east-west communications, surface application context, and validate whether security controls and tools have the visibility they require.

That independent evidence becomes even more important as security becomes increasingly automated. AI-driven systems need trusted data and context to inform and validate their decisions.

4. Economics Require Better Signals

AI-driven traffic growth also compounds an existing challenge for IT and security teams: telemetry overload. Logs, events, traces, alerts, flows, packets, and other data compete for processing, storage, analyst attention, and budget.

Collecting everything and sending it everywhere is neither economically nor operationally sustainable.

Organizations need to improve the quality of what reaches downstream platforms by filtering unnecessary traffic, eliminating duplication, enriching telemetry with application context, and transforming network traffic into higher-value telemetry before it reaches security, observability, data, and AI platforms.

The economics matter, too. As organizations invest more in AI infrastructure while telemetry ingestion, storage, processing, and tool costs continue to grow, a common foundation for network-derived telemetry can reduce redundant processing and help multiple downstream platforms access the data they need more efficiently.

Turning Network Traffic into Application Intelligence

The network has always provided a rich record of how applications, infrastructure, users, and data interact. As AI increases both the volume and complexity of those interactions, the ability to turn network traffic into useful application intelligence becomes more valuable.

The Gigamon Deep Observability Pipeline transforms network traffic into trusted network-derived telemetry, including packets, flows, and application metadata, giving security, observability, data, and AI platforms greater context while helping organizations manage the volume and cost of data sent downstream.

Our research reflects this evolution. Deep observability is growing as organizations contend with more traffic, greater infrastructure complexity, an expanding attack surface, and mounting pressure to extract better intelligence without allowing telemetry costs and infrastructure to scale at the same rate.

With 51 percent market share, Gigamon remains the category leader as the market expands. More broadly, the market’s trajectory points to deep observability becoming an increasingly important part of the infrastructure organizations will need to secure, manage, and optimize in the Hybrid AI era.

CONTINUE THE DISCUSSION

People are talking about this in the Gigamon Community’s AI Exchange group.

Share your thoughts today