The investment patterns of telecom infrastructure and AI look like a mismatch. PWC points out, “Telecoms’ capex intensity is easing after a decade of 5G and fibre build…capex will lag revenue growth through 2029 as many markets approach the end of their 5G non-standalone (NSA) and first‑wave fibre cycles.”
At the same time, “AI infrastructure is accelerating into a multiyear super‑cycle. Capital is flowing into hyperscale campuses, power‑hungry AI clusters, and adjacencies such as grid interconnects and onsite power generation. Data centres are now framed as “backbone” infrastructure for the digital economy” and spending is many, many billions.
PwC continues, “Overlay these contrasting cycles with today’s geopolitical fracturing and advancing sovereignty agenda, and what emerges is a once‑in‑a‑generation opportunity for legacy carriers.”
These opportunities are two-fold. The first is AI for networks which means integrating AI into networks and telcos’ systems to move towards higher levels of automation to gain business agility. This is closely related to networks for AI – that is building and offering enterprise customers access to AI infrastructure and systems instead of running their own.
The Briefing looks at operators’ attitudes and progress toward these opportunities, the drivers, what’s holding them back and what needs to happen next.
Battered from lack of monetisation from massive investment in 5G and the overall failure of Open RAN, how keen are operators to seize the AI infra opportunity? Or will it only appeal to a subset of operators?
There are big challenges to adapting and adopting technologies that are evolving so fast – and tremendous benefits to be had too. We consider what being an AI-native telco will look like by 2030 and the major milestones to getting there.