Enterprise connectivity is moving beyond the task of keeping circuits live.
Recent developments have pointed to a more ambitious model: networks that can be requested, provisioned, observed and increasingly troubleshot through AI-enabled workflows.
Equinix set the tone on 2 September with the launch of Fabric One at its inaugural Equinix Horizon event.
The offer – a managed connectivity service designed to link enterprise, cloud and AI infrastructure – is not simply faster interconnection.
Customers will be able to specify requirements through a portal, APIs, automation workflows, agents or natural-language prompts, while the platform manages routing, cloud connectivity, encryption, resiliency and failover.
“For decades, enterprise networks have been built one connection at a time for each partner and provider they depend on," said Chris Audie, Chief Product Officer at Equinix.
"That approach doesn't scale in a world of distributed AI that demands dynamic, flexible and real-time connectivity.”
“Equinix Fabric One will reduce the complexity that's slowing enterprises down by automatically managing connectivity across distributed architectures, based on the customer's intent.”
Fabric One is expected to enter beta later this year, before planned North American general availability in 2027, with AWS and Google Cloud as initial integration partners.
Network Intelligence Joins Service Workflows
Cisco ThousandEyes published a 3 September update on how its already generally available Model Context Protocol server can bring network intelligence into AI-agent workflows.
The company demonstrated how customers and partners can use the MCP server to build workflows that connect ThousandEyes with ServiceNow, Slack and other services.
The software, which reached GA in January 2026, is a meaningful crossover for service management.
A service desk agent dealing with a user issue could access network-path data without immediately switching to a separate operations toolset. The opportunity is a faster route from an incident report to root-cause evidence. But the risk is that AI accelerates poor decisions as readily as good ones, making governance, validation and clear human accountability essential.
ITSM Agents Get an Audit Trail
That concern is also reflected in ITSM provider Xurrent’s latest update. Its late-August release added an audit log for AI-agent runs, recording triggers, skills used, outcomes and actions.
It also introduced tools to review AI guardrail rewrites and retain classification rationale as an internal ticket note.
These may be relatively modest product changes, but they point to a more mature phase of ITSM automation. Organisations are increasingly asking not only whether an agent resolved a request, but how it reached its decision.
Performance Proof Takes Priority
Meanwhile, SmartCIC’s 4 September consolidation under the XVOLV name offers another indication of changing enterprise expectations.
The company has brought together its global connectivity-delivery business and network-intelligence operation, arguing that availability metrics alone no longer show whether a network is supporting the business.
“For two decades the test of a network was whether the circuit was live. That is no longer the question that matters,” said Toby Forman, Chief Executive of XVOLV.
“What matters is whether the applications the business depends on actually perform at every site.”
Taken together, the announcements tell a story of the continued evolution of the field.
AI is making networks more central to enterprise operations, while raising the bar for automation, observability and provable performance across both connectivity and service management.
