The more consequential detail is the architectural shift. ANA is replacing a model built around separate dedicated connections for individual systems with a centralized network-as-a-service control point, according to the Newsbytes report. That can remove a major operational bottleneck for an airline whose systems need to work across offices, airports, cloud providers and regions without waiting for a carrier circuit procurement every time an application team needs another connection.
ANA has not published a matching announcement in its publicly searchable English press-release archive, and Equinix’s global newsroom did not surface a release naming ANA at publication time. The 80% provisioning improvement, expected tenfold data growth and target of a 30% five-year total-cost-of-ownership reduction should therefore be treated as figures reported by Newsbytes and attributed to the deployment partners, rather than independently audited results.
What ANA has actually changed
Equinix Fabric is a software-defined interconnection service: a customer with infrastructure in an Equinix facility can create and manage private virtual connections to cloud providers, partners or other Equinix locations through a portal, APIs and infrastructure-as-code tools such as Terraform. Equinix’s own documentation describes Fabric as a way to establish direct, low-latency connections between distributed infrastructure and services, including AWS, Azure and Google Cloud.
For ANA, that is a practical answer to the old enterprise networking pattern: each new cloud adoption or regional expansion triggers a separate process for ordering and implementing connectivity. The central Fabric hub lets the airline make a change at the virtual-connection layer rather than rebuilding its connectivity design around every individual application and cloud tenant.
That does not mean ANA has eliminated its physical network. The Newsbytes account says Fabric avoids the need for separate physical circuits for each cloud platform, which is true in the narrow sense of the cloud interconnects. But Equinix’s technical documentation is explicit that Fabric rides on underlying physical infrastructure, ports, carriers and redundant routes. ANA still needs physical presence and access at the relevant Equinix locations, plus last-mile or wide-area connectivity to its own sites, airport systems and other facilities.
The savings and speed claims apply primarily to logical cloud connectivity, not to every part of an airline’s network estate. A new airport, branch or remote operational location can still involve local carrier lead times, hardware, contracts and regulatory requirements. The hub reduces one of the hardest layers to scale; it does not make physical network deployment disappear.
The 80% figure is plausible, but its boundary matters
Newsbytes reports that ANA reduced network provisioning time by about 80%, from several months to several weeks. That comparison makes sense if the previous baseline included procurement, cross-connect work, provider coordination and testing for dedicated circuits, while the new process lets ANA activate virtual circuits over an existing Equinix port.
Equinix markets Fabric as a service that can provision direct cloud connectivity in minutes. ANA’s reported result of several weeks is therefore not a sign that the platform itself takes weeks to create a virtual connection. Instead, it suggests the airline’s real process still includes enterprise steps outside the Fabric portal: network design, access approvals, change management, cloud-account configuration, security review, application testing and operational sign-off.
That distinction is useful for IT leaders looking at the case study. A self-service interconnection platform can collapse the carrier-order portion of a project, but it will not automatically fix internal approval queues or an application team that has not designed for multicloud routing, identity and observability.
ANA’s reported use of private links also deserves a similarly precise reading. Keeping traffic off the public internet can provide more predictable paths and reduce exposure to internet-facing attacks, but a private connection is not a complete security architecture. The published description does not identify the security products or controls ANA uses for segmentation, firewalling, encryption, identity, intrusion detection, key management or incident response.
Equinix Fabric can centralize connectivity management. It does not, by itself, prove that ANA has centralized every security policy across every cloud and on-premises system. That missing implementation detail is especially material for an airline connecting operational and customer-facing applications.
Why the change reaches beyond cloud networking
ANA’s public business materials show a carrier planning continued fleet and route growth, including additional Boeing 787-9 deliveries beginning in August 2026 and an ANA Group fleet expansion toward roughly 330 aircraft. More aircraft, routes and cargo operations translate into more distributed data sources: passenger and booking transactions, airport operations, flight planning, maintenance data, crew workflows, partner integrations and customer communications.
The airline’s reported expectation of a tenfold increase in global data volume is broad and has not been independently quantified. Yet the basic capacity problem is clear: once applications and data sets are split between more than one cloud, the network becomes a dependency of application delivery rather than a utility purchased once and left alone.
A centralized interconnection hub can simplify several recurring workflows:
- Development teams can request connectivity to approved cloud environments without redesigning the physical network around each project.
- Operations teams can establish more consistent routes among cloud platforms, regional systems and shared services.
- Network teams can govern connection inventories and bandwidth changes from one service rather than through a collection of carrier portals and point-to-point contracts.
- Architecture teams can avoid forcing every workload into one hyperscaler merely because that was the cloud already connected to the corporate network.
The last point may be ANA’s largest strategic benefit. A company that has paid for bespoke networking around one platform can become reluctant to use another, even where a different cloud offers better regional availability, services or commercial terms. Equinix’s neutral positioning gives ANA another option: it can use multiple providers without making each cloud decision a separate network construction project.
The cost target will be harder to verify than the provisioning gain
ANA is targeting a 30% reduction in total cost of ownership over five years, according to Newsbytes. That is a target, not a confirmed saving, and it bundles costs that are rarely disclosed in public: circuit commitments, colocation, cloud on-ramp charges, data egress, hardware refreshes, support contracts, staffing and the cost of downtime or delayed projects.
There is a credible route to savings. A centralized virtual interconnection service can reduce redundant circuits, shorten manual engineering work and let a company right-size bandwidth rather than buy capacity for a long procurement cycle. Direct private cloud connectivity may also let an enterprise manage traffic patterns that would otherwise generate more expensive internet transit or cloud egress charges.
But there is a counterweight. Equinix Fabric has its own recurring port and virtual-connection costs, and a multicloud design can increase operating complexity if teams duplicate tools, data, identity controls and skills across providers. An airline will only reach the five-year figure if it retires enough legacy circuits and avoids building a new overlay of operational work on top of them.
No published breakdown identifies which legacy links ANA will retire, what percentage of traffic will move through Fabric, whether all three major clouds are already live, or how the target accounts for cloud data-transfer charges. Those omissions do not undermine the project, but they prevent readers from treating the 30% number as a comparable benchmark for their own environment.
The operational test is resilience, not provision speed
ANA’s systems have different tolerance for delay. A development sandbox can wait for a change window; reservations, passenger check-in, boarding and flight operations cannot. Centralization makes the network easier to manage, but it also raises the importance of redundancy, failure isolation and disaster-recovery design around the hub.
Equinix says Fabric’s underlying network uses multiple carriers, diverse routes and redundant circuits. That is useful infrastructure resilience, but ANA’s end-to-end result will depend on how it has designed ports, metro diversity, cloud on-ramps, routing failover and the paths between its business locations and Equinix. A private interconnect is only as resilient as the least redundant segment attached to it.
For Windows-heavy enterprise environments, the change has a familiar consequence. Active Directory and Entra ID integrations, Microsoft 365-adjacent services, Azure-hosted workloads, endpoint-management platforms, virtual desktop infrastructure and SQL Server or Windows Server applications can all become easier to place across regions when the network team can activate governed private paths quickly. They do not become simpler to secure or recover merely because the connection request took weeks instead of months.
ANA’s move is therefore best understood as an operational modernization project, not an AI announcement wearing a network label. The airline may use the capacity for analytics, automation and future AI workloads, as Newsbytes reports, but the immediate value is more basic: engineers can connect approved systems faster without ordering a new dedicated circuit for every cloud decision. The reported five-year cost target will only become meaningful once ANA discloses how much of its legacy network it actually retires.