In the Future of Voice AI series of interviews, I ask three questions to my guests:
- What problems do you currently see in Enterprise Voice AI?
- How does your company solve these problems?
- What solutions do you envision in the next 5 years?This episode’s guest is Geoff Burbridge, Principal + Founder at Human Edge Advisory.
Geoff Burbridge is Principal + Founder of Human Edge Advisory, where he helps executive teams, boards, and founders transform organizations at the intersection of people, process, and technology. With more than 30 years of executive leadership experience at Bank of America, USAA, Truist, and Capital One, Geoff advises organizations on AI readiness, operating model design, and enterprise transformation, helping leaders balance innovation with the security, governance, and trust required by today's enterprises. He is a keynote speaker, executive moderator, and board advisor who believes that as technology continues to evolve, the greatest competitive advantage will always be the Human Edge.
Recap Video
Takeaways
AI’s biggest contact center impact may be stopping fraud, not replacing agents.
The first AI strategy will often be wrong because the market, not the roadmap, reveals the real use case.
BPOs may emerge as AI winners, not casualties, because they can apply automation across massive service operations.
The industry has already retreated from “replace the agent” to “make the agent better.”
AI adoption will stall when companies automate tasks but ignore how the workforce itself must change.
The real AI risk is not failure, but destroying customer confidence before the technology earns their trust.
Fraudsters face don’t the same rules as enterprises, giving attackers a speed advantage that regulation alone cannot solve.
Voice fraud is becoming an AI-versus-AI fight that humans cannot manage at machine speed.
Most contact centers already have the signals to stop fraud but cannot connect them fast enough to act.
Authentication failure is weak protection when attackers can rapidly change devices, numbers, identities, and voices.
The customer call companies want to eliminate may be the exact interaction that stops a major fraud loss.
Human trust makes voice service powerful, but it also makes voice one of the easiest channels to exploit.
Deepfake scams succeed by creating urgency before the victim has time to question what sounds real.
The best fraud systems will not just detect known attacks; they will identify new attack patterns as they form.
Agentic AI is more valuable for absorbing demand spikes than replacing every customer conversation.
The best automation targets are processes that already reach the right outcome almost every time.
Repeatedly failed journeys need more human judgment, not another layer of automation.
AI coaching can scale practice without the pressure and resistance that make traditional role-play fail.
The intersectionality of AI is where its value compounds, connecting fraud signals, customer context, automation, and human judgment in real time.
AI’s real power comes from connecting fraud signals, customer context, agent judgment, and automation in the same moment.
Contact centers overcomplicate service when customers fundamentally only care about four things: answering the phone, being nice, resolution, and ease.
In Geoff’s experience, authentic service earns trust because customers believe people, not scripts.
