Why AI will help strengthen wholesale voice, not reinvent it
Author: Harry Huang, Product Manager
“Just call me”.
When chatbots fail, when your Wi-Fi is playing up, or when you need to verify that email really came from your bank...what do you turn to? A straightforward phone call.
There’s a reason people defer to voice time and again, despite the wealth of alternatives. It is universally adopted, always reliable, and completely auditable.
It remains one of the most sustainable and enduring core services on an operator’s roster. In fact, traditional wholesale market is expected to grow at a rate of 10.6% CAGR between now and 2035. Despite these facts, industry rhetoric continues to tout that wholesale voice is ‘dying’.
As a result, organizations are often tempted to reinvent their voice models, and the increasing accessibility of AI will likely accelerate this approach.
However, in the case of voice: less is more.
A margin optimization model
The reason voice was labelled as a failing service was not because it was not only losing market share to VoIP and OTT, but because it became all about margins and a race to the bottom.
Operators could no longer charge premiums to supplement declining revenue. They had to focus on cutting costs, optimizing efficiencies, and maximizing margins – which made profitability an ongoing concern.
AI to strengthen voice, not replace it
AI and automation present big possibilities for transforming voice. However, using these technologies to make incremental evolutions that improve margins and strengthen the service’s core proposition will be key to its ongoing success.
Instead of rewriting voice models, organizations can use AI to reduce risk, and reinforce voice’s existing reputation for integrity and trust.
For example:
Operators can use the incredibly analysis capabilities AI offers to drive forward smarter decision making.
AI can detect early warning signs of fraud, identify trends across thousands or even millions of records, and flag suspicious behavior or numbers. AI can build the fraudulent number databases, while operators retain authority on what to block.
Similarly, AI offers predictive visibility over potential billing shortfalls or delayed payments that are often only discovered retroactively, helping organizations avoid costly disputes and gain tighter margin control.
Finally, everything from routing to rating can be optimized and semi-automated to reduce the usual anomalies and misconfigurations that can squeeze margins.
In these scenarios, humans remain in control of the decisions.
'Less is more’ builds continued resilience
Voice isn't dying. It has simply become defined by resilience, trust and longevity rather than disruptive growth. This may look like stagnation – but it really demonstrates an ability to evolve without abandoning its core qualities.
The real risk to voice now is doing too much to change it. Large-scale change could introduce complexity, cost, and risk – while doing nothing allows inefficiencies to grow.
Finding the balance between traditional models and transformation can strengthen what voice does best. By improving fraud detection, billing accuracy, and operational visibility, AI helps trim long-standing inefficiencies, without disrupting relationships or removing human judgement from the equation.
This ‘less is more’ approach will not just protect margins. It deepens wholesale voice’s role as one of the most traceable, accountable, and reliable forms of global communication.
