Telcos: Being Right vs. Being Paid

The telecom market is not a place to prove you are right. It is a place to make money. Being right about the technology and being profitable on the investment are separate skills; a lesson worth remembering as the AI gold rush triggers another enormous infrastructure cycle.

Telecom has spent the last decade proving that Open RAN, private 5G, edge computing, network APIs, 5G standalone, and network slicing can all work. The problem is that working and making money are not the same thing.

The pattern is always the same. A new technology appears, standards get written, vendors publish enormous TAMs, consultants draw a hockey stick, operators spend billions, and a few years later the technology is still alive while the original revenue forecast has quietly disappeared.

That is key these days because telecom is standing next to an even larger pile of capital labeled AI. Telcos are again being shown huge opportunities around inference, sovereign AI, GPU-as-a-service, AI-RAN, and edge infrastructure. Some of them may be real. But before spending another few billion proving they were technologically right, operators should ask a much simpler question:

If we are right, who actually gets paid?

Private 5G is a good place to start because, technically, it worked. By the end of 2025, roughly 6,500 private LTE and 5G networks were deployed worldwide, up from 4,700 a year earlier. The problem is that the entire market was worth only about $2.4 billion, while forecasts had spent years pointing toward opportunities of $80 billion, $100 billion, and even $150 billion. Nobody lied about the technology. The factories got connected, and the networks got deployed. The money simply forgot to show up at the scale promised.

Network APIs are following a similar, concerning track. Telcos can now expose identity, fraud, location, and network capabilities through standardized APIs, which is genuinely useful. Actual operator API revenue in 2025, however, was roughly $284 million, while forecasts continue to point toward a few billion by 2030, although earlier industry estimates talked about hundreds of billions in broader value creation. Apparently, the easiest part of creating a $100 billion telecom market remains putting $100 billion in the title of the consulting report.