Voice infrastructure planning has to account for what AI will do to call volume, and that assumption doesn't just shape a forecast; it shapes capacity planning, carrier agreements, and how hard you fight for voice budget over the next two years. Increasingly, the answer looks less like deflection and more like growth.
New research from AVOXI and Metrigy's 2025 State of International Voice for the Contact Center report points to a clear trend taking shape. 82% of enterprises say AI is already driving more traffic to their voice channel, or expect it to soon. Not fewer calls, handled more efficiently. More calls, period. And this isn't a prediction about some future state. 81% of organizations have already enabled their voice environments with AI, so they are speaking from experience rather than speculating.
Why automation increases volume instead of reducing it
The answer is basic economics. AI lowers the cost of handling a call. When the cost of something falls, businesses don't usually bank all of the savings. A company that can staff, route, and resolve calls more cheaply doesn't shrink its voice channel. It widens who gets access to it.
Voice was never actually unpopular. It was rationed. Live agents are expensive, so companies pushed customers toward chat, email, and self-service portals to protect their budgets, not because those channels won on their merits. Cheap, AI-supported voice removes the reason for rationing. The gate that kept callers out starts to open, and volume climbs.
This lines up with something contact center operators and IT directors already know from experience: customers reach for voice specifically when the stakes are highest. As AVOXI CEO Barbara Dondiego wrote in this year's report, “Voice remains the most effective way to resolve urgent and complex issues, and the preferred method for both customers and agents.” Voice was never a weak channel. It was an expensive one. AI just made it less expensive.
You can already see the shift in how companies are using AI in voice today. The leading use cases are self-service (71%), optimized call routing (67%), and real-time agent assist (65%). Every one of them does the same thing: it increases the amount of voice traffic an organization can absorb without adding headcount. That's not a deflection engine. That's a volume engine.
What rising volume means for four planning decisions
A plan built on the shrinking-channel assumption encounters trouble in four areas.
Capacity. If your carrier agreements, number inventory, and routing capacity were sized around last year's flat-volume assumption, this is where that assumption is first tested. As AI takes hold, historical capacity trends might not be a good predictor of the future, so it's important to monitor what’s happening across your voice channel more actively and frequently.
Budget. Voice shifts from a line item you defend to one you actively grow. That's a different conversation with finance — better to walk in with it than have volume force the conversation later.
Architecture. 84% of companies now have at least some portion of their contact center in the cloud, but 65% still maintain on-premises systems in some form. Hybrid environments can absorb this at flat call volume. It gets expensive fast once volume rises and calls have to route through legacy hardware that wasn't built for growth.
Vendor strategy. 96% of organizations already run more than one voice provider. Every one of those relationships is a separate integration point that must reliably stream voice to both AI systems and human agents, and fragmented vendor relationships make it harder to do consistently. Consolidating around a strategic voice partner means having one clean, AI-ready path to route calls, rather than several inconsistent ones.
None of this is theoretical for the 82% already seeing it. The infrastructure question is whether your capacity, budget, and architecture are built for the direction the data points, and whether your vendor strategy is built for an AI-mediated channel, not just a bigger one
The market is already rethinking voice partners
The shift is showing up in vendor strategy, too. Within their unified communications environment, 79% of organizations expect their voice needs to change within the next one to two years. For 40% of them, that means moving to a new preferred provider. Nearly a third, 34%, expect to move specifically to a specialty global voice provider, and they're doing it for two reasons: greater control (60%) and a better ability to leverage AI (55%).
That's worth reading carefully. Companies aren't shopping for a new voice vendor because the old one broke. They're re-evaluating voice as a strategic infrastructure they need more control over, not a utility they renew on autopilot. When AI turns a channel into a growth driver, “good enough” stops being good enough.
The planning question that actually matters
The right question isn't “how much will AI reduce our call volume?” It's “what happens to our infrastructure if volume goes up 20% and many of those calls is machine-mediated?” Capacity, budget, architecture, and vendor strategy all answer that question differently than they answer the shrinking-channel version.
You Might Also Be Interested in
Need Help Getting US Phone Numbers?
We're here to help! Contact us today so we can help find the right business phone number for you.
