AI Takes the Call Centre by Storm — But Human Agents Aren’t Out of the Game Yet
The call‑centre floor might be changing faster than many of us realise. Advanced artificial intelligence (AI) systems are rapidly being deployed in customer‑service roles, yet the transition is far from smooth, and human workers remain a critical part of the picture.
The Big Picture
Companies are increasingly using intelligent agents—AI-driven chatbots that can resolve customer issues without human intervention. Analysts predict that by as early as 2029, up to 80 % of common customer‑service queries could be handled autonomously by AI.
One global services firm invested hundreds of millions of pounds in these systems; over nine-in-ten customers opted into interacting with AI chatbots. Yet many deployments still fall short. Studies suggest only about 20 % of companies achieve their AI-powered customer service targets.
Performance vs Reality
AI offers clear advantages: fast responses, 24/7 availability, data-driven answers, and scalability. One company reported cutting US$100 million in support costs after deploying AI.
However, AI struggles with nuance. In one widely-publicized case, a delivery‑company chatbot insisted a parcel had been delivered incorrectly, confusing customers. Analysts warn that AI agents often hallucinate—confidently providing false information—or fail at tasks requiring empathy or complex judgment.
Real-World Case: Netwealth’s AI Rollout
Company: Netwealth (Australia) — a retail investment and retirement platform.
Challenge: Handling increasing customer inquiries while maintaining service quality and managing costs.
Implementation:
- Introduced an AI-powered support layer: intelligent ticket routing, chatbots, voice assistants, and AI-assisted live calls.
- Built a comprehensive knowledge base from FAQs, transcripts, and policies.
- Trained staff to handle escalations and view AI as a productivity tool, not a threat.
Outcomes:
- Agents reachable within 60 seconds on average.
- Nearly 99% of calls resolved in a single touch after AI rollout.
- Efficiency gains included fewer transfers, faster responses, and more focus on complex queries.
- Customer satisfaction held steady or improved, showing AI did not degrade experience.
Lessons Learned:
- Good data is foundational: AI needs high-quality, up-to-date knowledge.
- Hybrid model works best: AI handles routine queries; humans manage complexity.
- Agent buy-in is critical: Staff support drives adoption.
- Continuous monitoring: Metrics like AHT, CSAT, and transfer rates ensure AI effectiveness.
- Customer experience is paramount: Frictionless handoff to humans is key.
What This Means for Call Centres & Workers
- Shifting jobs: Staff can focus on high-value work like escalations or retention.
- Training data is gold: Without strong data, AI underperforms.
- Regulatory pressure rising: Disclosure requirements for AI interactions are increasing in Europe and the US.
- Hybrid models dominate: AI plus human oversight is the practical route today.
- Consumer trust hinges on quality: A bad AI experience can erode brand confidence fast.
Takeaway
AI isn’t magic, but it can now handle a large chunk of customer‑service work autonomously. For organisations with strong data and human-in-the-loop oversight, the upside is tangible. For those lacking fundamentals, AI risks wasted investment and frustrated customers.
Human judgment, empathy, and complex problem-solving remain indispensable. Call centres aren’t vanishing—they’re evolving.
Glossary
- Agent (intelligent agent): Software that autonomously performs tasks, such as handling customer queries.
- Knowledge base: A repository of structured information that AI draws on to answer questions.
- Hallucination (AI): When an AI confidently provides false or fabricated information.
- First-line support: Initial customer contact handling standard queries.
- Human-in-the-loop: Humans supervise or intervene in AI processes.
Source: BBC
