Huawei predicts autonomous AI agents will drive over 90% of global AI token traffic by 2035, requiring significant increases in computing power and new security systems. The Chinese tech giant's forecast contrasts with some US calls to slow AI development, emphasizing China's push to accelerate adoption while managing risks.
Huawei's projection of autonomous agents dominating AI traffic underscores the immense future demand for computing power and advanced AI infrastructure, potentially shaping global technology development and competition, particularly between China and the U.S.
China's Huawei expects autonomous AI agents to dominate global AI traffic by 2035, accounting for over 90% of token consumption. The tech giant's "Intelligent World 2035" report anticipates a massive increase in computing power and the development of new systems to ensure the security and controllability of these independent software entities.
The forecast contrasts with some discussions in the U.S. technology sector about potentially slowing the development of advanced AI capabilities due to concerns about rogue agent incidents and the difficulty of controlling future systems. Huawei's report, however, does not advocate for a slowdown, instead assuming a dramatic expansion of agent capabilities and deployment. It identifies security and privacy protection for autonomous agents as crucial among ten necessary technologies.
This approach highlights a divergence in AI development strategies, with parts of the U.S. industry considering coordinated limits while China generally aims to accelerate AI adoption across its economy, concurrently building technical standards to address emerging risks. Huawei believes integrating AI into the physical world is essential for achieving artificial general intelligence (AGI), defined as systems capable of performing a wide range of tasks at or beyond human levels.
Building on its earlier projections, Huawei anticipates approximately 900 billion AI agents could be active globally by 2035, a figure roughly 100 times the projected human population. The company is also leveraging AI agents for security, having launched an autonomous security-operations platform in April designed to detect and respond to cyber threats. Furthermore, Huawei has proposed AI firewalls to safeguard enterprise AI systems.
Guo Ping, chairman of Huawei's supervisory board, recently remarked that AI represents the company's "biggest opportunity," expressing a desire for its computing infrastructure to become China's equivalent to U.S. chipmaker Nvidia.