Huawei forecast that autonomous AI agents will account for more than 90 percent of global artificial intelligence (AI) token traffic by 2035. It also expected demand for computing and network infrastructure to surge as AI agents spread.

On Sept. 16, blockchain media outlet Cryptopolitan reported that Huawei made the forecast in a research report released that day titled "Intelligent World 2035: Turning Vision Into Action". The report was unveiled at an online event held under the theme "Act now. Create the future." Huawei also released the "Global Digitalization and Intelligence Index 2026" at the same event.

In AI systems, tokens are units of data processed as AI models take in and generate information. Huawei analysed that token use would rise sharply because AI agents, unlike existing application-centered AI usage, continuously perceive their surroundings, carry out reasoning and decision-making, and call external tools.

Huawei therefore expected global annual token consumption in 2035 to be 100,000 times higher than now, with AI agents accounting for more than 90 percent. It said an agent-centred internet environment would expand computing and network demand in a different way from the existing application-centred internet.

Huawei also forecast that about 900 billion AI agents would be operating worldwide in 2035. That would be about 100 times the world population expected at that time. Huawei presented 10 core technology directions to support AI agents expected to spread on such a large scale.

Huawei proposed expanding computing cluster scale by 100 times while cutting the cost per agent task to one-thousandth. It said this would require moving beyond simple chip performance improvements and evolving into "SuperPoD" computing systems. Key tasks included storage and memory, an agent operating system (OS), intelligent connectivity, chip design, intelligent driving, power and cooling for AI data centres, and security and privacy protection.

Huawei's outlook contrasts with debates in the United States over the pace of AI development. Some U.S. researchers and technology industry executives have argued that advanced AI development should slow, citing hard-to-predict AI agent behaviour, autonomous control of equipment and the possibility of securing resources. By contrast, Huawei has focused on developing technologies to secure safety and controllability on the premise that AI agents will spread.

Huawei is already developing security technologies that use AI agents. In April, it launched an autonomous platform for security operations centres (SOC), designed to detect and respond to cyber threats using monitoring and response agents.

Executives also set out positions on expanding AI infrastructure. Guo Ping (궈핑), Huawei's chairman of the board of directors, called AI Huawei's biggest opportunity and said Huawei's computing infrastructure should play a role in China similar to Nvidia. David Wang (David Wang), Huawei's vice chairman and rotating chairman, said agent-based AI would be a key element of future change, and that while the direction is clear, concrete execution is needed to make it reality.

Huawei also sees AI expanding beyond software into the physical world as a key step toward artificial general intelligence (AGI). It explained that AI interacting with the physical world must be able to perceive, judge and act in its environment to evolve into AGI capable of performing complex tasks.

The Global Digitalization and Intelligence Index 2026 released alongside it analysed the digital and intelligentisation level of 90 countries and forecast that AI would create more than $27 trillion in cumulative economic value over the next five years.