89% of Telecoms Are Raising AI Budgets as Network Automation Overtakes Customer Service

By Hector Herrera | April 12, 2026 | Telecom

Eighty-nine percent of global telecom operators plan to increase AI spending over the next 12 months—up from 65% last year—and for the first time, network automation has overtaken customer experience as the top deployment priority. The numbers come from a new NVIDIA survey of global carriers, and they mark a strategic pivot in how telecoms think about AI's role in their business.

What Happened

NVIDIA's 2026 telecom AI survey found 89% of operators planning AI budget increases, with network automation now ranked above customer service as the top use case for both investment and ROI impact. Separately, SoundHound AI announced a partnership with Associated Carrier Group this week to deploy agentic AI across telecom customer service operations—a deal that reflects the customer-facing AI investment that continues even as network automation moves to the top.

Carriers are also more optimistic about 6G than historical telecom rollout timelines would predict: 77% expect faster-than-historical 6G deployment.

Context

Telecom operators have invested in AI for customer service applications for several years—chatbots, call center automation, churn prediction. These are visible to consumers and have produced measurable cost reductions in customer operations. They're also table stakes now: every major carrier has them in some form.

The shift to network automation represents a deeper architectural change. Telecom networks—the physical and software infrastructure that routes calls, data, and connections—have traditionally required constant human operator intervention to monitor performance, identify faults, route around failures, and optimize capacity allocation. AI-native network management, often described as self-healing or autonomous networks, aims to perform most of these functions without continuous human direction.

This matters for 6G in particular. The performance specifications for 6G networks—much lower latency, much higher reliability, much more granular service guarantees—are difficult to achieve with human-managed operations. AI-native management is not an option for 6G; it's a prerequisite.

Details

The jump from 65% to 89% of carriers increasing AI budgets in a single year is unusually sharp. It suggests the industry crossed a threshold in 2025—either in demonstrated ROI from early deployments, in competitive pressure from carriers that moved earlier, or in confidence that the technology is mature enough for critical network infrastructure.

Network automation use cases that carriers are actively deploying include:

AI-native RAN (Radio Access Networks): The software that manages radio frequencies and connections in mobile networks. AI-driven RAN can optimize spectrum allocation in real time, adjusting to traffic patterns and interference conditions faster than human operators can.

Self-healing networks: Systems that detect faults, identify root causes, and route around failures automatically—reducing the outage duration and eliminating the need for 24/7 human monitoring of routine network events.

Predictive capacity management: AI systems that anticipate traffic demand—a major event, a network spike, infrastructure failure—and pre-position capacity before congestion or failure occurs.

The SoundHound AI partnership with Associated Carrier Group targets the customer service side specifically. Associated Carrier Group represents independent wireless carriers—smaller regional and rural operators. The deal brings agentic AI customer service capabilities to carriers that couldn't build these systems independently, suggesting the technology is now accessible enough to reach beyond the tier-1 operators who moved first.

Impact

For telecom customers: The most immediate consumer benefit is reliability. AI self-healing networks mean faster recovery from outages and fewer dropped connections. The performance improvements from AI-native RAN show up as better signal quality and more consistent speeds, particularly in congested environments.

For telecom workforces: Network operations centers—the teams that monitor network performance and respond to faults—are the most directly affected by network automation. These are skilled technical roles. The automation wave doesn't eliminate network engineering; it changes what network engineers do, shifting from monitoring and reactive troubleshooting toward planning, governance, and handling edge cases that AI systems can't resolve.

For enterprise buyers of telecom services: AI-native network management enables service level agreements that weren't previously feasible—not just uptime guarantees, but latency guarantees, quality guarantees, and rapid response commitments. Enterprises buying connectivity for latency-sensitive applications (manufacturing automation, autonomous vehicle infrastructure, real-time financial systems) should push for SLA terms that reflect AI-native network capabilities.

For equipment vendors: Nokia, Ericsson, Samsung, and Huawei are all building AI-native network management into their RAN and core network products. The carriers increasing AI budgets are buying these systems. Equipment vendors that can demonstrate real network performance improvements from their AI-native products will take share from those that can't.

What to Watch

The 77% of carriers expecting faster-than-historical 6G deployment is the most forward-looking data point in the survey. Telecom has a long history of technology timelines slipping—3G, 4G, and 5G all deployed slower than early projections. The carrier optimism about 6G speed-to-market is based on the premise that AI-native network management reduces deployment complexity. That premise will be tested as 6G trials move from lab to real-world pilots in the next 18 to 24 months.

Also watch for regulatory activity around AI-managed critical infrastructure. Telecom networks are essential services in every major economy. When AI systems are making autonomous decisions about network routing, fault response, and capacity allocation, questions about liability, audit, and human oversight become regulatory issues. That framework doesn't exist yet.

Hector Herrera covers telecom and AI for NexChron.