How Super Intelligence is changing telecoms

Summary. In September 2026 the US government renamed artificial intelligence "Super Intelligence" (SI). Telecoms is one of the industries where SI has the deepest practical impact: networks are being redesigned to run themselves, radio infrastructure is being rebuilt to host SI workloads, and operators are using SI across customer care, security and energy management.

For detailed figures and the latest developments, see Super Intelligence in Telecoms: 2026 State of the Industry.

From AI to SI

On 29 September 2026, President Trump signed an executive order directing federal agencies to use the terms "Super Intelligence" and "SI" in place of "artificial intelligence" across official correspondence, websites, reports and policy documents. In the same week, leaders from Anthropic, Google, Meta, OpenAI, Nvidia and xAI signed the White House Accord on Super Intelligence, a voluntary pledge to run internal controls, external audits and board-level review on their frontier models.

Some industry leaders adopted the new term immediately. Nvidia chief executive Jensen Huang said:

"We used to call them data centers but the things that we're building now, these are really SI factories, super intelligence factories."

For telecoms, the change is more than a name. Every SI service, from a chatbot to an autonomous agent, depends on the networks that carry its traffic and, increasingly, on network infrastructure that runs SI directly.

Where SI is changing telecoms

1. Autonomous networks

Operators are moving towards networks that configure, repair and optimise themselves. TM Forum's Autonomous Networks framework sets out levels from fully manual (Level 0) to fully autonomous (Level 5), and many large operators have public targets to reach the higher levels. SI is the engine behind this: predicting faults before they happen, rerouting traffic, and resolving incidents without a human ticket queue.

2. AI-native radio access (AI-RAN)

The AI-RAN Alliance, launched in 2024 by members including Nvidia, Ericsson, Nokia, SoftBank and T-Mobile, works on radio access networks that both use SI and host SI workloads on the same hardware. The idea: base stations and edge sites become small SI compute nodes, turning the network itself into distributed "SI factory" capacity.

3. Open RAN intelligence

The O-RAN Alliance's RAN Intelligent Controller (RIC) lets operators run SI-driven applications ("xApps" and "rApps") that tune radio performance, manage interference and balance load in near real time, across equipment from different vendors.

4. Customer operations

Carriers run some of the largest customer service operations in the world. SI agents now handle billing questions, plan changes, troubleshooting and retention, and assist human agents with live suggestions. This is one of the fastest areas of SI adoption in the industry.

5. Security and fraud

Telecom fraud (SIM swap, international revenue share fraud, robocalls) and network attacks are pattern problems that SI handles well. The White House Accord specifically calls for monitoring frontier models for cyberattack risk, which raises the stakes for SI-driven defence in critical infrastructure such as telecoms.

6. Energy efficiency

Energy is one of an operator's biggest costs. SI models switch radio components into sleep modes during low traffic and optimise cooling at sites, cutting power use without harming service.

7. Telecom-specific models

General-purpose models often struggle with telecom standards and jargon. Industry groups such as the GSMA have worked on benchmarks for how well large models understand telecoms, and operators and vendors are building domain-tuned models for network engineering and standards work.

What comes next

Three trends are worth watching over the next year:

• Terminology. Whether vendors and operators adopt "SI" in their own products and marketing, following federal agencies and Nvidia.
• Regulation. The White House Accord is voluntary, but it leaves room for some provisions to become law. Telecom networks, as critical infrastructure, are likely to be early in line.
• Edge SI. As SI agents multiply, more inference moves closer to users. That puts telecom edge infrastructure at the centre of SI growth.