Telecom firms increasing AI spending and automation, NVIDIA survey reveals

NVIDIA telecom survey finds companies increasing AI spending after seeing gains in revenue, efficiency, and productivity.

Telecom companies are increasing their investment in AI as the technology begins to reshape how networks operate and how services are delivered. Survey results suggest the shift is already affecting both revenue and efficiency, while also speeding up plans for AI-driven network systems.

NVIDIA's fourth annual State of AI in Telecommunications report shows that most telecom firms now see clear financial value from AI. About 90% of respondents said the technology is helping grow annual revenue while cutting costs. Interest in generative AI is also rising, with 60% of organizations using or testing it, up from 49% the previous year.

Investment plans reflect that momentum. Nearly nine in ten telecom companies expect to increase AI spending in 2026, compared with 65% in last year's survey. More than a third expect their budgets to rise by over 10%.

Sebastian Barros, managing director of Circles, said the industry is undergoing a "seismic shift" driven by AI, as operators begin to rethink their broader role. He explained that providers are moving beyond simply transmitting data and toward supporting intelligence across local, regulated infrastructure — a transition he described as the path from traditional telcos to "AI infrastructure companies" working close to the network edge.

Automation becomes the main focus for AI investment

AI efforts in telecom are increasingly centered on automating networks rather than only improving customer experience. Around 65% of telecom operators said AI is already driving automation across their infrastructure.

The goal is to build autonomous networks — systems designed to configure, repair, and optimize themselves with minimal human input. Most companies say they remain in the early stages of this process, operating at levels 1 to 3 of autonomy under TM Forum definitions. Advances in generative AI and agent-based systems are expected to push networks closer to full autonomy.

Barros said autonomous networks can deliver quick returns by removing manual work from repetitive and reactive tasks, pointing to areas such as energy management, fault prediction, configuration correction, and capacity planning as the fastest to benefit.

Industry analysts also see automation as one of the clearest financial wins for telecom AI projects. Chetan Sharma, CEO of Chetan Sharma Consulting, said autonomous systems tend to produce faster returns because they reduce outages, power use, and the need for manual fixes. He added that agent-based AI helps speed this up by coordinating decisions across systems in real time.

AI-native wireless systems may arrive before 6G

Telecom companies are also adjusting network design to support AI processing closer to end users. Spending on edge computing is rising as firms prepare distributed infrastructure that can run AI tasks locally instead of relying only on centralized data centers.

This shift connects to work on AI-native radio access networks and future wireless standards. About 77% of survey respondents believe AI-native networks could launch before full 6G deployment, suggesting architectural changes may arrive sooner than the next generation of wireless itself.

Key reasons for these investments include improving spectrum efficiency, strengthening radio network performance for edge AI applications, and accelerating wireless research and development.

Generative AI leads telecom workloads

Among different AI workloads, generative AI now ranks first in telecom adoption. Sixty percent of respondents listed it as a primary workload, with usage especially high among large companies with more than 10,000 employees, where adoption reaches 80%.

Data processing and analytics follow at 58%, while agent-based AI systems stand at 48%.

Telecom operators report using AI across many functions. About half said they are exploring or deploying AI-native network technologies for AI-RAN or 6G research. Other uses include conversational AI tools at 50% and predictive analytics at 47%.

Looking at operational priorities, 54% of organizations now use AI for network planning, deployment, and optimization — a rise of 17 percentage points from last year. Customer service improvement comes next at 46%, followed by internal workflow improvements in departments such as IT, HR, and finance at 43%. Wireless research and development accounts for 28%.

Productivity gains spread across telecom operations

Beyond revenue and network performance, telecom firms also report internal efficiency gains from AI tools. Nearly all survey respondents said AI is improving employee productivity, with about a quarter describing the improvement as major.

These gains come from both generative AI tools and newer agent-based systems used across operations, from back-office work to network management.

Sharma said generative AI has already brought quick productivity improvements, but noted that agent-based AI is where telecom companies start to see deeper structural returns. Autonomous agents, he said, can operate across networks, IT systems, and customer processes, turning insights into actions without waiting for human input.