AI in customer service: balancing technology & the human touch
With National Customer Service Week in focus and CX Day 2025 on the horizon, support leaders are rethinking how AI fits into the contact centre. The mission is simple: use AI to reduce friction and increase empathy, so customers feel understood and supported.
AI as an enhancer
Chris Angus, Vice President for CPaaS and CX Expansion, EMEA at 8x8, notes the conversation has moved past hype to outcomes customers feel in every interaction: speed, clarity, and empathy. "As we mark CX Day 2025, it's never been more important to reflect on what 'great customer experience' really means, and how artificial intelligence is reshaping it... customers aren't impressed by buzzwords. They're impressed by seamless service, speed, empathy, and a sense that the brands they interact with understand them. That's where AI can be transformative - if we use it right."
He's clear that AI isn't a silver bullet. It's a toolkit. "Agents don't fear AI. They're crying out for it... AI helps remove the mundane, such as repetitive queries, manual logging, basic triage, so agents are freed to focus on what humans do best: empathise, problem-solve, and build trust." Done well, this shift turns agents into proactive advisors and improves satisfaction, engagement, and burnout-all of which customers notice.
The continuing role of humans
Lisa Orford, Global Vice President for Contact Centre at 8x8, stresses that technology should enhance the human element, not replace it. "Customer experience is the bedrock for companies to build trust, but that experience isn't just up until the point of sale. Organisations need unshakeable loyalty to stay ahead in this competitive market, and that comes from exceptional experiences across the entire buying journey."
She points to research showing companies that treat service as a value centre see 3.5x revenue growth. Frontline insight is a growth engine-especially when paired with smart tech. "No AI tool on the planet can or should replace the human element... Chatbots may solve problems, but humans are the ones people remember."
Practical plays you can run now
- Route with intent: Use AI for intent detection, triage, and smart routing. Set clear handoff rules and latency targets so high-stakes issues reach humans fast.
- Agent assist, not agent replace: Surface knowledge, similar cases, and next best actions during live conversations. Keep knowledge bases governed and up to date.
- Automate the admin: Auto-summarise calls, classify topics, and log dispositions to cut after-call work and reduce handle time.
- Quality at scale: Use conversation analytics for targeted QA, sentiment, and outcome tagging. Create short coaching clips and close the loop with agents.
- Be proactive: Spot repetitive issues and publish fixes in-product, in help centres, and on status pages to prevent avoidable contacts.
- Measure what matters: Track CSAT, FCR, AHT, containment rate, escalation rate, agent effort, time to proficiency, and cost per contact.
- Start small, prove value: Pilot a low-risk queue, A/B test bot flows, share results with frontline teams, and iterate weekly.
Guardrails that keep trust intact
- Clear disclosure: Tell customers when they're speaking with an automated agent and make human handoff easy.
- Fairness and quality: Review training data, test across segments, and monitor failure modes. Keep humans in the loop for sensitive use cases.
- Data discipline: Limit retention, redact PII, and apply strong access controls. Document prompts, models, and decision logic.
- Agent-first change: Train supervisors and agents, align incentives, and make AI performance visible so teams can improve flow by flow.
What this means for support leaders
Great service comes from people who are backed by smart systems. The winning approach combines purposeful AI deployment with empowered, empathetic teams.
Focus your next quarter on two tracks: remove repetitive work with automation, and raise the quality of human interactions with better tools, coaching, and context. Keep customer promises tight and measurable.
Events and resources
Explore CX Day with the Customer Experience Professionals Association for ideas and community support: CXPA. For UK teams marking National Customer Service Week, see the Institute of Customer Service: NCSW.
If you're upskilling your team on practical AI for support roles, browse curated training by job function here: Complete AI Training.