AI puts new demands on telecom networks: Ericsson

AI applications are driving telecom operators to restructure their networks to prioritize data uploads over downloads, according to Ericsson India's managing director Nitin Bansal. As more data is generated by AI-enabled devices such as glasses and sent to the cloud, the demand for a robust uplink side of mobile networks is increasing, Bansal told ET AI at the India Mobile Congress 2026.

Bansal referred to this shift as "networks for AI," emphasizing the need for network capabilities tailored to AI applications. In addition, Ericsson is exploring "AI for networks," where the technology is utilized to enhance network performance and efficiency.

For telecom operators, these advancements could result in improved spectrum utilization, reduced energy consumption, and the ability to support a larger number of users without a proportional increase in network resources. AI can also aid in lowering the power consumption of network equipment, ultimately decreasing operating costs for operators.

The distinction between "AI for networks" and "networks for AI" is crucial as operators seek ways to monetize investments in AI-ready infrastructure. Enhanced uplink capabilities for AI-enabled devices could lead to better user experiences and increased consumption of connectivity.

Ericsson's India manufacturing efforts, which have produced over 15,000 antennas and 25 million components since July 2025, are geared towards supporting the transition towards more intelligent and programmable networks. These enhancements include improving coverage and capacity, strengthening uplink performance, using spectrum more efficiently, and reducing network energy consumption.