83% of UK Consumers Prefer Human Agents, But 30% Would Accept AI for Lower Prices

UK customers want human support. AI gets a pass when it saves money

New 8x8 research makes the trade-off clear: 83% of UK consumers prefer speaking to a human for customer service. Only 4% prefer chatbots.

But cost changes the equation. 30% would accept AI if it lowers prices. For support leaders, that means a human-first service with AI working behind the scenes - and clear, tangible savings when automation is customer-facing.

Key stats at a glance

• 83% prefer human agents; 4% prefer virtual agents or chatbots.
• 30% would choose AI if it reduces prices (men 34%, women 28%).
• Regional price sensitivity: London 45% willing to accept AI for lower costs; Scotland 20% (lowest).
• By age: 16-24 show 7% AI preference and 17% no preference; human-first still dominates across all groups.
• Scotland and Wales: 90% pro-human for urgent matters; Northern Ireland shows ~11% AI preference.

Quote to align your strategy

"Customers still value a human touch - but they're also pragmatic. People will choose AI if it saves them time or money… The best contact centers today are using AI as a teammate, not as a substitute." - Chris Angus, VP for CPaaS and CX Expansion at 8x8

What this means for customer support leaders

Human for urgency, AI for leverage. Use AI to shorten handle time, prep agents, and clean up ops. Be explicit about the value trade-off when you ask customers to interact with automation.

• Urgent = human-first: Fast access to an agent, clear escalation paths, and zero dead-ends.
• Simple + low stakes = guided automation: Let AI handle verification, status checks, and routing - with one-tap escape to a human.
• Price transparency: If you offer an "AI-assisted" tier, attach a visible discount or guaranteed faster resolution to earn consent.

Regional and demographic signals you can act on

• London: Highest openness to AI for lower prices (45%). Test AI-forward offers with clear savings.
• Scotland: Lowest price-driven acceptance (20%). Lead with premium human support.
• 16-24: Slightly more open to AI, but still prefer humans. Keep the "agent in the loop."

Where AI delivers real value today

• Agent co-pilot: Real-time suggestions, knowledge surfacing, form-filling, and after-call summaries.
• Smart triage: Intent detection, routing, and pre-qualification with immediate agent handoff when confidence drops.
• Analytics: Spot trending issues, capacity gaps, and training opportunities from interaction data.
• Workflow automation: Identity checks, record updates, order lookups, and follow-ups run in the background.

Decision rules: automate or route to a human

• Automate if low risk, low emotion, and high repeatability (password reset, delivery status, appointment scheduling).
• Hybrid if a task is structured but context-heavy; AI prepares context and drafts, agent confirms.
• Human if urgent, high value, regulated, or emotionally charged - or if automation confidence is below a defined threshold.

Voice cloning: proceed with care

Replica voice split opinions: 43% opposed, 28% supportive. Younger adults (25-34) show 48% support; only 12% of 55+ are positive.

Use clear consent, opt-outs, and labeling. Keep replica voices out of urgent or sensitive flows. Test in non-critical use cases first.

Implementation checklist

• NLU accuracy: Track intent confidence and fallbacks. Auto-escalate below a confidence threshold.
• Systems access: Connect CRM, orders, knowledge, and policy data to avoid "sorry, I can't do that" loops.
• Training data: Continuously feed resolved cases and edge scenarios; tune for bias and gaps.
• Human oversight: Monitor CSAT, containment, AHT, and recontact rate. Review transcripts where customers exit automation.
• Controls: One-click agent transfer, clear wait-time expectations, and transparent pricing for AI-tier experiences.

Metrics to manage

• Customer: CSAT by channel, effort score, first-contact resolution, complaint rate post-automation.
• Ops: AHT, agent occupancy, containment with satisfaction, escalation rate from AI to human.
• Finance: Cost per contact, discount payback for AI tiers, churn impact.

90-day rollout plan

• Weeks 1-2: Map top 20 intents by volume and emotion; define "never automate" rules for urgent flows.
• Weeks 3-6: Launch co-pilot for agents and background automation for repetitive admin. Start with one channel.
• Weeks 7-10: Add smart triage and limited self-service for two low-risk intents with instant human escape.
• Weeks 11-13: A/B test AI-assisted pricing or fast-lane options in London; keep Scotland human-first.

Context snapshot

• Survey: 2,000 UK adults, Feb 24-26, 2025; results released Oct 16, 2025.
• Industry trend: Major CX platforms are adding AI agents and co-pilots; early adopters report positive ROI, especially in support.

About the survey

• Nationally representative UK sample, adults 16+.
• Strong preference for human agents on urgent issues across all regions.
• Modest increase in AI acceptance for non-urgent matters, but human preference remains high.

Next steps

• Codify your "urgent = human" policy and publish it.
• Deploy AI where it helps agents work faster, not where it frustrates customers.
• Test AI-for-savings offers in price-sensitive regions; keep a premium human option everywhere.

Helpful resource

Build team skills for human-in-the-loop support and AI co-pilots with focused learning tracks: Courses by job.