Telcos are embracing AI whole-heartedly, according to Nvidia
In sum – what we know:
- State of AI in Telecommunications – 90% of survey respondents state AI is helping to increase annual revenue and drive down costs.
- Investment surge – 89% of telecom operators plan to increase AI spending in 2026, with network automation as a major driver.
- Network evolution – 77% of respondents expect to see AI-native networks launch prior to the deployment of 6G.
Nvidia’s fourth annual “State of AI in Telecommunications” survey has been released, pulling from roughly 1,000 respondents. The takeaway is that operators are putting big money behind AI to drive business results, and the data reflects that across the board.
Obviously, this is survey data from a company that has skin in the game, and that that sells AI chips into the telecom market. That said, the trends showing up here, especially around investment growth and the pivot to network automation, track with broader industry momentum that’s genuinely hard to wave away.
Investment return
According to Nvidia’s report, 90% of respondents said AI is actively increasing their annual revenue while driving costs down. Telecom operators specifically reported an equally strong 90% positive impact across both revenue and cost metrics. The top use cases delivering ROI were autonomous networks at 50%, better customer service at 41%, and internal process optimization at 33%.
Also interesting is where the money is going next. 89% of respondents said they’re increasing their AI budgets over the next 12 months, which is a big jump from the 65% who said the same thing a year ago. Among those planning increases, 35% expect budgets to grow by more than 10%. “There is a seismic shift underway in the telecom industry driven by AI,” said Sebastian Barros, managing director of Circles, a Singapore-based telecommunications provider. “Communication service providers are converging on a new realization. Their role in society extends beyond moving bits across networks toward moving intelligence across local and regulated infrastructure.”
Autonomous networks
Network automation has now overtaken customer experience as the top priority for AI investment, deployment, and ROI impact, according to Nvidia. The survey found that 65% of telecom operators said AI is the primary force behind their network automation efforts.
There’s still a significant gap between ambition and reality, though. Around 88% of organizations place themselves between Level 1 and Level 3 on the TM Forum’s autonomy framework, which maxes out at Level 5 for fully autonomous operations. “Autonomous networks are delivering return on investment faster than any other AI use case because they directly reduce outages, energy consumption and manual intervention,” said Chetan Sharma, CEO of Chetan Sharma Consulting. The expectation is that generative AI and agentic AI will help accelerate progress toward higher autonomy levels, though nobody seems willing to put a firm date on reaching Level 5.
On a related front, edge computing investment is ramping up as operators push to bring AI inferencing closer to end users through distributed infrastructure. The key investment drivers include using AI to boost spectral efficiency, improving radio access network performance for edge AI workloads, and accelerating 6G R&D. Perhaps most notably, 77% of respondents expect AI-native networks to go live before 6G formally deploys — which suggests the industry isn’t waiting around for the next-generation standard before fundamentally rethinking how wireless networks are built and operated.
Productivity and agentic AI
AI’s impact is showing up in day-to-day productivity, too, Nvidia says. Nearly every survey respondent reported that AI is making employees more productive, with 26% citing significant improvements. Generative AI adoption keeps climbing as well, with 60% of organizations now either using or evaluating the technology, up from 49% in 2024.
Open source is turning out to be a big part of how telcos are building their AI strategies. Some 89% of respondents flagged open source models and software as important, which makes sense given the industry’s long-standing allergy to vendor lock-in and its deeply collaborative approach to standards development.
The bigger shift, though, is the move from generative AI toward agentic AI — systems that don’t just respond to prompts but autonomously coordinate decisions across multiple domains in real time. “Generative AI delivered fast productivity gains, but agentic AI is where telecoms begin to see structural ROI,” Sharma said. “Autonomous agents can act across networks, IT and customer journeys, turning insights into decisions without human delay.”
