AI workloads require a total structural reset in networks, says Nokia

Nokia says artificial intelligence is no longer a niche add‑on but a scaled, mobile‑native workload that is reshaping how networks must be built and run. Speaking at MWC Barcelona, Nokia chief Justin Hotard argued that operators can’t rely on traditional, SLA‑driven capacity planning if they want to support AI reliably. Instead, he said, networks need deterministic, programmable architectures that tightly link data centers, transport networks, and edge sites so AI traffic can move with predictable latency and performance.

According to Nokia, the volume and behavior of AI traffic will drive this structural reset. As AI processing spreads from central clouds toward the edge, networks will have to coordinate compute, storage, and connectivity as one system rather than as separate domains. That shift, Nokia says, will demand more automation in the network, more granular control over traffic flows, and infrastructure designed from the outset around AI workloads rather than retrofitted for them.