We’re packing up our bags and heading to the airport after a successful EC, but our biggest insights aren’t staying in Vegas. We’re sharing what we saw and heard on the floor and in speaking sessions at this year’s biggest gathering of enterprise IT and infrastructure leaders.
1. Hybrid is here to stay
We’re seeing a lot of enterprises that have telecoms half in the cloud, and half on-prem. If you’d asked us five years ago if we’d see this permanent hybrid state, we’d have told you no. It felt like a one-way rush to the cloud was underway. But we’re seeing organizations’ need for redundant systems making them hesitant to put it all in the cloud.
Enterprise contact center leaders are shy of marketing pitches claiming the cloud is ready for full enterprise hosting, because they don’t necessarily have the baked-in reliability of an on-prem PBX. Not only that, many enterprises are still working with inherited systems which may be under contract for on-prem systems. Plus, those legacy systems have familiarity baked in, which can make switching hard.
The truth is, you really can have your cake and eat it too. The value is realized when you divest from the capital expenses required to upkeep on-prem infrastructure and when you can easily take advantage of AI. And, as you decide which platform(s) your enterprise will utilize across communications, bringing a carrier that has the right global footprint and built-in network reliability and redundancy will ensure you maintain 99.99% uptime.
In summary, cloud migration is a continuing trend, but it’s not a one-size-fits-all destination yet.
2. Resilience is king
Resilience and failover are key considerations for all enterprises, but especially for those in highly regulated industries such as finance and healthcare. We’ve heard C-suite leaders looking for ways to maintain data sovereignty and communications continuity that are redundant and disaster-ready. It’s one of the reasons some enterprises are staying hybrid. But the right cloud providers provide enterprises with the resilience they need through smart redundancy.
For example, at Bandwidth we’re taking our years of experience in providing 5x redundant voice networks, and creating new resilience for the AI era with features like:
- Independent failover from the CCaaS platform: Because we route calls to your AI platform first, your AI experience remains available even if the CCaaS platform is degraded or offline, and vice versa.
- Geo‑redundant call routing: Your calls are automatically distributed and rerouted to a geographically diverse regions to protect against localized outages.
- AI provider redundancy: You can configure primary and secondary AI endpoints. If the primary AI system becomes unavailable, we’ll automatically redirect traffic to a backup AI instance or a fallback IVR.
3. AI is here, and it requires interoperability
Our Director of Solutions and Go-to-Market, Zach Kunkel, attended David Evans’ session titled “Honey, I Blew Up the Contact Center.” With a title like that there were guaranteed to be good takeaways, and Zach summarized the key highlight as “technology is moving too rapidly for any one distinct system to keep up with the needs of a Global 2000 enterprise serving their customers at the highest level. Interoperability and orchestration are key.”
The advancements in AI are fast and furious, and enterprises need systems that can handle frequent updates. Just because a system worked for pre-AI voice, doesn’t mean it’ll have the security and features needed to connect with agents today. And, just as important, it doesn’t mean you’ll be set up for whatever comes next in AI.
The solution is maintaining a flexible operating system that lets you swap out integrations as needed. And this aligns perfectly with our philosophy at Bandwidth: CIOs deserve freedom of choice to build and evolve their preferred CX stack without vendor lock-in. This is exactly what our “Bring Your Own AI” approach lets you do in Maestro!
4. AI is in pilot mode
Enterprises in 2026 are past the AI hype cycle, but many are discovering that moving from hype to real value is harder than anticipated. While nearly every organization has launched at least one AI pilot, whether through Claude, Microsoft Copilot, ChatGPT, or custom-built tools, the gap between early experimentation and enterprise-wide transformation remains wide.
The core challenges aren’t just technical: organizations are wrestling with how to build the business case, navigate data integration hurdles, secure end-user buy-in, and balance the build vs. buy decision, all while keeping security and compliance intact.
The enterprises finding real success share a common playbook: they started with targeted pilots, identified clear use cases tied to collaboration and productivity gains, and then systematically scaled what worked. But scaling isn’t easy — cost pressures, organizational resistance, and underwhelming results from poorly scoped deployments have slowed momentum for a significant portion of the market. The conversation is shifting from “should we invest in AI?” to “how do we make our AI investments actually pay off?”
5. Trust is gold in 2026
As fraud levels continue to rise, the importance of trust has never been higher. And we heard this loud and clear at Enterprise Connect, where one panel even included “Why Security Must Be an Obsession in the AI Future” in its title.
Companies have been adding more scrutiny and investment in security over the last few years, and we think it’ll continue to be an important trend for the foreseeable future. That’s why Bandwidth is leading the charge on brand and call integrity, by providing solutions for both outbound and inbound trust.
On the outbound side, our enterprise customers come to us for:
- Number Reputation Management
- Identity Authentication, which is a pre-call API to verify calls at the network level, preventing bad actors from spoofing your brand
- Identity Presentation, which displays your brand name (coming in Q2)
For inbound calling, we’ve been leading the security charge for years with:
- Call Verification for spoof and ANI identification
- Our Pindrop integration, which uses voice biometrics to authenticate calls and detect deepfakes
We don’t stop there, though. Our advanced Insights like Anomaly Detection for Voice use an AI/ML model to look for deviations in your specific calling patterns, which may indicate fraudulent activity.
We’ve backed up our tradeshow booth, but we’ll be reflecting on our Enterprise Connect conversations all year. Because at Bandwidth, we’re here to support enterprises who are ready to defy compromise, take command of their communications, and build secure, customer-friendly experiences that harness the latest AI technology while maintaining reliable uptime.
Next up on our events calendar are back to back London events: Cavell Summit Europe on March 25th and Genesys Xperience London on April 28th and 29th. We’d love to see you there!
