AI boom will need bigger telecom ‘pipes’; spectrum still a hurdle: COAI chief

India’s push to scale artificial intelligence (AI) will require greater telecom network capacity as the industry seeks more spectrum for rising data traffic, lieutenant general Dr SP Kochhar, director general of the Cellular Operators Association of India (COAI), told ET AI during the India Mobile Congress (IMC) 2026.

Kochhar said telecom networks will require thicker data pipelines and fresh investments in infrastructure to handle a growing volume of both uploads and downloads.

“The pipes we need will have to be much thicker,” he said, adding that telecom operators would have to keep expanding network capacity and using bandwidth more efficiently amid AI acceleration.


Unlike conventional internet usage, where download speeds typically exceed upload speeds, AI applications could drive a more balanced demand for uploading and downloading data, Kochhar said. So, networks must evolve alongside investments in data centres and AI infrastructure.

However, spectrum availability remains a concern for the industry. COAI has been pushing the government to allocate more spectrum in the 6 GHz band, arguing that operators need more capacity to support 5G and prepare for future network requirements.

“Our requirement for 5G spectrum is 2 GHz, but we have received only 700 MHz,” Kochhar said, adding that the industry is seeking another 1,200 MHz to meet what it considers its minimum requirement.

Also read: AI can help factories find bottlenecks before production stops: Tata Communications’ Praveen Arora

The demand comes as India looks to strengthen its position in next-generation telecom technologies, including 6G. “When we say that we want to become leaders, it is not only in deployment. We are talking about 6G standards, patents, applications, and then comes deployment,” Kochar said.

While spectrum requirements are expected to rise with 6G, technologies such as edge computing, edge storage and network slicing could help manage some of the additional demand by processing data closer to where it is needed, he said.

Monetisation, a concern

For telecom operators, expanding capacity comes with a financial burden. Kochhar said the industry remains concerned about the scale of investments required, particularly as operators continue to look for viable ways to monetise 5G.

“Globally, we still don’t have a clear monetisation case or use case for 5G,” he said, suggesting that new applications emerging with 6G could create investment opportunities.

Kochhar also called on the government to reduce spectrum and other acquisition costs, and give operators greater flexibility in determining how they price and offer services.

“We are investing billions of rupees in networks, so give us the opportunity to derive benefits from the market,” he said, arguing that regulation should provide an overall framework without being overly prescriptive.


Also read: AI applications will force telecom networks to rethink uplink, spectrum and power: Ericsson India MD

AI & cybersecurity challenges

Kochhar warned that expansion of AI-driven traffic will raise cybersecurity concerns as the technology could strengthen the capabilities of both network defenders and cybercriminals.

“AI helps us, but it also helps criminals,” he said, adding that telecom operators would have to continuously adapt their defences as threats evolve.

He called for regulatory frameworks to keep pace with technological developments and greater accountability across the digital ecosystem, including OTT platforms and other services.

He also highlighted data sovereignty as a growing consideration with India investing in data centres and AI infrastructure. While international cloud services may be cheaper than building domestic infrastructure, he said, sovereign cloud capacity would give India greater control over its data.

“We have to strengthen and Indianise our databases, then build LLMs and SLMs on them,” he said, arguing that India needs AI models suited to its own requirements.

Also read: AI agents don't take coffee breaks, so networks must scale up: Ciena's Gautam Billa