What's Happening?
A recent study by Cisco projects that artificial intelligence (AI) will become a significant driver of internet traffic growth over the next decade. The report, titled 'AI Impact on Wide Area Networks,' suggests that as consumer adoption of AI reaches
near-universal levels, AI-driven network traffic could increase by approximately 6.6 times by the mid-2030s. This growth is expected to account for about 63% of additional network traffic compared to non-AI scenarios. Currently, AI inference traffic is minimal compared to other categories like video streaming, but it is anticipated to represent 25% of total network traffic by 2035. The study highlights that AI traffic differs from typical web traffic, with longer flows and higher upstream capacity requirements. Cisco notes that AI agents, which perform a variety of tasks, are emerging as 'power users,' generating significantly more traffic per task than traditional web transactions.
Why It's Important?
The anticipated surge in AI-driven internet traffic has significant implications for network infrastructure and service providers. As AI becomes a dominant force in internet traffic, network operators will need to adapt by implementing 'flow-aware network and security systems' to manage the increased and altered traffic patterns. This shift will likely necessitate upgrades to existing infrastructure, such as enhancing upstream capacity and reducing latency. For broadband operators, this means investing in technologies like DOCSIS networks to achieve symmetrical speeds and low latency. The changes in traffic patterns could also influence how service providers differentiate their offerings, potentially leading to new competitive dynamics in the broadband market. As AI continues to reshape internet traffic, stakeholders across the industry will need to adjust their strategies to accommodate these changes.
What's Next?
As AI traffic continues to grow, network operators are expected to roll out new technologies and infrastructure upgrades to handle the increased demand. This includes implementing low-latency capabilities and expanding upstream capacity. Companies like Comcast and Charter Communications are already making strides in this area by introducing low-latency features in their networks. Additionally, as AI agents become more prevalent, the need for advanced network management tools will become more critical. These developments will likely lead to further innovation in network technology and could drive new business models for service providers. The industry will need to closely monitor AI traffic trends and adjust their infrastructure and service offerings accordingly to remain competitive.
