This global survey reveals a brutal truth about AI in CX
This Global Survey Reveals a Brutal Truth About AI in Customer Service
Entrepreneur’s new survey of 6,000 consumers finds that the “human” part of human-in-the-loop AI-powered customer experiences is critical. Some key findings:
- 31% would hang up immediately upon being connected to an AI agent
- 79% would choose a business offering a human receptionist over an otherwise comparable competitor
- 57% say their trust in a brand decreases when it relies mainly on AI for service
- 82% have actively asked to speak to a human instead of AI.
The data lands as enterprises are racing to reduce human headcount in support. We’re used to automated customer service; IVR and chatbots are nothing new. The key is to use AI to do a better job than the previous iterations of the technology. Speed and cost efficiency matter, but these numbers are a warning that the pendulum can swing too far. The most forward-thinking CX leaders are thinking of how to use AI to drive meaningful business outcomes that drive a better return for the business, not just shave off costs. AI behind the scenes (and in front of customers as a way to speed up service) has the power to help customer experience teams deliver better experiences that drive loyalty and revenue. Don’t settle for the simplest applications of AI.
Should Your Subscription Business Use Auto-Renew?
New HBR research challenges a core CX assumption. While auto-renewal boosts short-term retention, it suppresses trial sign-ups so sharply that opting for auto-cancel ultimately produces more paying subscribers and higher-quality customers over the long run. The subscription economy now totals nearly $500 billion annually and is projected to exceed $1.5 trillion by 2033, with the average American holding 7–12 paid subscriptions simultaneously. If your retention strategy leans heavily on inertia, you may be trading long-term customer quality for short-term numbers.
Exceptional Customer Service Wins Loyalty — But Companies Are Still Missing the Mark
A new Verint survey finds 51% of consumers say companies are failing to meet their service expectations — up from 46% the prior year — while 42% say their own expectations have risen in the past 12 months. That's a widening gap. Four in five consumers will make another purchase after an exceptional experience, and the same proportion will switch to a competitor after a poor one. The good news: 79% acknowledge AI provides real benefits like faster resolution, and over two-thirds say they'd accept automation if it actually solved their issue.
Ultimately, the bar is rising faster than CX orgs are clearing it. It’s not because the tools aren't there, but because experiences aren't being designed around consistent and scalable outcomes.
Tune In
Is AI actually transforming customer experience, or is the reality far messier than the hype? In this episode, Tue Søttrup, CEO of SmartRole, separates boardroom mandates from ground-level truth. Tue shares why most enterprise AI implementations fail, what agents actually need to feel supported through the transition, and what the best CX organizations will look like five years from now.
In this episode:
- Why CX leaders are getting AI wrong — and how to flip the approach
- The hidden cost of automating easy tickets
- How to earn agent buy-in without the fear
- The service recovery paradox and why humans still win
- What proactive, AI-powered CX actually looks like
AI Headlines
1) Academic researchers studying emerging agent-to-agent collaboration networks found that autonomous AI systems are increasingly sharing reusable workflows and coordinating tasks with minimal human intervention. The next phase of AI may involve networks of collaborating agents, not just isolated models.
2) Robinhood now lets AI agents trade stocks. AI Agents in Robinhood can connect to its MCP server to analyze and execute stocks, plus have a dedicated wallet to ensure spending limits. It shows how AI agents will handle more transactions on a customer’s behalf.
3) Former OpenAI CTO Mira Murati outlined a vision for AI systems designed to keep humans actively involved rather than fully automate decision-making. Her startup, Thinking Machines Lab, is developing interaction-first AI models focused on responsiveness and collaboration. The debate over AI autonomy versus human oversight is becoming one of the defining strategic divides in the industry.
4) Anthropic is briefing global financial regulators on cybersecurity risks tied to its advanced Claude Mythos model, which demonstrated the ability to uncover previously unknown vulnerabilities in IT systems. The industry is entering a phase where frontier AI capabilities may outpace existing cybersecurity defenses and governance frameworks.
5) Insurance leaders at the Innovators USA conference described AI enabling concierge-level service: proactively detecting totaled vehicles, arranging replacements, alerting customers to market events. Air Canada's chatbot lawsuit was cited as a reminder that governance must keep pace with ambition.
