Human Agents Pay the Hidden Operational Cost of AI in Customer Service

Only 28% say their systems can seamlessly transfer customer context across channels, exposing how fragmented many CX environments still are. Is AI creating additional layers of supervision and correction?

Customer service organisations have spent years expanding digital channels, introducing automation tools, and layering AI into support operations designed around speed and efficiency. But as AI adoption accelerates, many companies are discovering that automation alone does not necessarily reduce operational complexity.

In many environments, AI is exposing how fragmented customer service systems still are.

The issue is no longer whether organisations are investing in AI-powered customer service. Most already are. The larger challenge is whether the systems surrounding those tools are connected enough for automation to function effectively across the customer journey. Without unified customer context, synchronised channels, and coordinated workflows, AI can create additional layers of supervision and correction rather than removing operational effort altogether.

That friction is becoming increasingly visible across customer interactions. Customers move between chatbots, apps, websites, voice support, and human representatives, expecting continuity across every touchpoint. Instead, many still encounter repeated questions, disconnected conversations, inconsistent information, and fragmented service experiences that weaken trust and prolong resolution times.

These operational gaps are reflected in findings from Capgemini Research Institute’s Reimagining Customer Experience: Human-led, AI-powered report, which found that 83% of executives acknowledge their organisations struggle to provide seamless transitions between channels. At the same time, only 28% say their systems can successfully transfer customer context and conversation history across online and offline interactions.

The gap between AI ambition and operational readiness is becoming harder for organisations to ignore.

AI Is Increasing Oversight Alongside Automation

Much of the current conversation around AI in customer service focuses on efficiency gains, automation, and reduced workloads. The findings suggest the operational reality is more complicated.

AI systems are increasingly capable of handling repetitive interactions, summarising conversations, routing tickets, and supporting frontline operations. However, automation is also creating new forms of monitoring and correction work for customer service teams when systems lack coordination underneath.

Frontline customer service teams are increasingly expected to review AI-generated outputs, ensure accuracy, intervene when responses fail, and manage customer frustration when automated interactions break down. Instead of fully removing work, automation often redistributes it across different parts of the operation.

The report highlights how widespread these coordination issues have become. Around 63% of consumers say they frequently need to repeat issues to human representatives after already interacting with chatbots, while 72% of frontline employees report friction when navigating multiple systems to access customer information during support interactions.

Those failures are not simply customer experience issues. They point to deeper orchestration problems inside customer service environments where systems, channels, and customer records still operate in isolation from one another.

The consequences become more significant as organisations scale AI adoption further. AI systems rely heavily on shared context, connected workflows, and synchronised data environments to deliver interactions that feel seamless. When those conditions do not exist, automation can amplify fragmentation rather than reduce it.

AI Adoption is Moving Faster than Operational Coordination

Many organisations are now scaling AI capabilities faster than they are solving the infrastructure challenges underneath them.

The report found that 78% of organisations are shifting toward outcome-driven AI interactions, while 68% believe AI agents will eventually outperform traditional customer service channels in 

At the same time, only 23% say they currently operate with a comprehensive customer experience strategy spanning the entire customer journey. Another 40% cite the absence of clear roadmaps and defined KPIs as one of the biggest barriers slowing CX transformation efforts.

That imbalance is creating a growing disconnect between AI capability and operational execution.

Many organisations now possess advanced automation tools, but the environments in which those systems operate remain fragmented across departments, channels, and data sources. As a result, companies risk scaling inconsistent experiences faster rather than improving customer service quality in a meaningful way.

The same disconnect is becoming visible in personalisation efforts. While 66% of organisations believe they are successfully using customer data to personalise interactions across channels, 61% of consumers say brands still fail to deliver experiences that justify the amount of personal data being collected.

Human Resolution Still Shapes Customer Trust

Even as automation expands, the findings reinforce that customers continue to value experiences that feel responsive, contextual, and human during complex interactions.

The report found that 69% of consumers can forgive product-related issues, but not poor customer experiences. Another 72% expect organisations to make them feel valued during every interaction, while 77% want companies to listen more carefully to customer feedback and act on it.

This creates a balancing act that many organisations are still learning to manage. AI can improve speed and operational scale, but efficiency alone does not automatically build trust or customer loyalty.

The organisations most likely to succeed with AI in customer service will not necessarily be the ones deploying the most AI tools. They will more likely be the companies capable of building connected systems, coordinated workflows, and customer journeys that still feel consistent and human underneath the automation.