Swan.my.id - Jakarta, Huawei predicts that AI agents will account for more than 90% of global AI token traffic by 2035. The company expects this shift to create huge demand for computing power while also requiring new systems to keep autonomous software safe and controllable.
Huawei outlined the projection in its latest report, Intelligent World 2035: Turning Vision into Action. The report says AI token consumption will rise sharply over the next decade as AI agents take on more independent tasks.
Tokens are the units that AI models process when they receive or generate information. Unlike conventional software, AI agents can continuously observe their environment, reason through situations, make decisions, and call external tools or services without direct instructions for every step.
AI Agents Could Reshape Computing Demand
Huawei links the expected increase in token traffic to the growing autonomy of AI systems. As agents handle more complex tasks, they will need to process information repeatedly and interact with a wider range of applications and devices.
That pattern could place significant pressure on computing infrastructure. Data centers, processors, networks, and software platforms would all need to support larger volumes of AI activity. Consequently, the expansion of AI agents may affect how companies plan their technology investments over the next decade.
Huawei previously projected that around 900 billion AI agents could operate globally by 2035. The company estimated that figure would equal roughly 100 times the projected human population during that period.
Autonomy Raises Security and Control Concerns
The rapid development of AI agents has also raised concerns about safety and human control. Researchers and AI executives in the United States have discussed a possible slowdown in frontier AI development following incidents involving AI agents.
Some warnings focus on future systems that could operate hardware or obtain resources independently. Such capabilities could make the systems harder for people to monitor and control.
Huawei’s report does not call for a pause in AI development. Instead, the company expects AI capabilities and adoption to grow quickly. It lists security and privacy protection for autonomous agents among 10 technologies needed to support that growth.
Meanwhile, the debate reflects a broader difference in the way major technology markets approach AI development. Some parts of the US technology industry have discussed coordinated limits on the advancement of frontier AI. China, by contrast, generally continues to promote faster AI adoption across economic sectors while developing technical standards to address emerging risks.
Huawei Expands Its Own AI Security Efforts
Huawei also views the use of AI in the physical world as an important step toward artificial general intelligence, or AGI. The term generally refers to systems that can perform a broad range of tasks at a human level or beyond.
The company has already started applying AI agents to cybersecurity. In April, Huawei launched an autonomous security operations platform that uses agents to detect and respond to cyber threats.
In addition, Huawei has proposed an AI firewall to protect AI systems used by businesses. These initiatives show how the company is connecting the expansion of autonomous software with stronger security infrastructure.
Huawei Supervisory Board Chairman Guo Ping described AI as the company’s biggest opportunity in a statement to employees. He also said Huawei wants its computing infrastructure to become the Chinese equivalent of Nvidia, the US-based maker of AI chips.
With AI agents expected to handle more observation, reasoning, and decision-making, the next phase of development will involve more than model performance. Computing capacity, privacy safeguards, and reliable human oversight will also shape how quickly autonomous AI becomes part of everyday systems.
