AI in Customer Service: What Changes for Agents, When Humans Step In and How to Measure AI

• Question #1: Will AI Replace Customer Service Agents?
• Question #2: What Will the Customer Service Agent Role Look Like in the Future?
• Question #3: When Should a Human Handle Customer Service Instead of AI?
• Question #4: How Do You Performance-Manage AI in a Contact Center?

Vancouver, BC, Oct. 06, 2026 (GLOBE NEWSWIRE) -- As enterprise contact centers move AI past the pilot stage, their leaders are converging on the same short list of questions about where AI fits, what changes for human agents and how to manage it once it is live. TELUS Digital, a customer experience transformation partner that operates the full AI value chain with enterprises, including training, building, running and protecting AI, has published its answers to the four questions support leaders ask. The company's position across both contact center operations and enterprise AI shapes how it answers them. The goal is better outcomes for the business, the customer and the agent, delivered through whichever technology fits the use case, whether that is TELUS Digital's Fuel iX™ platform or a partner's.

Erin Walker, Global Vice President of CX AI, Business & Delivery at TELUS Digital, said, "The central theme behind each of these questions goes back to understanding how people and AI divide the work, and who is accountable for the result. AI should take the parts of a customer interaction that slow an agent down, so the agent can spend their judgment where it counts. Get that division right for each use case and you get a better experience for the customer and a better job for the agent at the same time."

KEY FACTS

• TELUS Digital is a customer experience transformation partner that operates the full AI value chain with enterprises, training, building, running and protecting AI through its products and a deep technology-partner ecosystem.
• TELUS Digital's approach centers on delivering better business, customer and agent outcomes through the technology that fits each use case, rather than a single tool.
• Fuel iX is TELUS Digital's proprietary enterprise AI platform and suite of products for clients to manage, monitor and maintain generative AI across the enterprise.

Will AI Replace Customer Service Agents?

AI is changing the kind of work agents do, not eliminating the need for them. As automation takes on the simple, repetitive interactions, it becomes possible to handle some contacts autonomously end to end, while in others, AI elevates the work agents do, allowing them to focus on the judgment calls, problem-solving and human connection that a customer needs most. The roles shift toward that higher-skill work rather than disappearing.

Different contexts call for different approaches. Some routine, well-defined interactions can be handled by automation, including agentic AI that completes a task end to end. Others need a person from the first moment. Many sit in between, where AI assembles context and drafts a path and the agent decides. Fraud-related and compliance-heavy interactions are their own category, where regulatory requirements and the cost of an error make human validation essential rather than optional. The work lies in matching each interaction type to the right approach instead of applying one answer across the board. In the deployments TELUS Digital works on, AI is used to enable agents rather than replace them. When AI helps an agent resolve an issue faster and more accurately, the result is a better experience for the customer, which is the actual goal.

What Will the Customer Service Agent Role Look Like in the Future?

As automation absorbs routine, well-documented requests, the agent's role concentrates on the interactions that need a person. These include emotionally charged conversations, ambiguous account issues and the escalations where a customer has exhausted self-service options and is still seeking resolution. They also include fraud and regulated-industry cases, where the agent has to confirm the system's output is correct before acting on it. That shift raises the skill profile of the role. Agents spend more of their time on complex, higher-value interactions and less on repetitive lookups, which changes how their performance is developed and measured. Speed to proficiency becomes a more meaningful measure than the raw volume of calls handled because the remaining work depends on skill rather than speed.

When Should a Human Handle Customer Service Instead of AI?

Routine, well-documented requests, such as order status updates, account changes and password resets, are suited to automation, but several situations point the other way. Complex or emotionally charged interactions, such as a billing dispute that needs judgment or a service failure that needs an urgent fix, are where customers want to work with a person. Instances involving identity checks, suspected fraudulent transactions and regulated disclosures require a human to validate what the automation surfaces. This is where customer services and regulatory compliance overlap.

High-value customers and high-stakes interactions warrant a concierge level of service that a human agent equipped with AI is better positioned to deliver. When a customer has worked through the available self-service tools and chatbots and is not getting anywhere, they need a live agent even if the issue is simple and the customer is calm. Effective customer experience design instantly handles routine requests through automation and makes sure a human agent is reachable the moment the automated path stops working with the context already assembled so the customer does not start over.

How Do You Performance-Manage AI in a Contact Center?

Managing an AI deployment is closer to workforce management than to installing software. The same discipline a contact center applies to staffing, scheduling and agent development, where it matches capacity to demand and coaching performance over time, applies to the AI in the operation. That means deciding which interactions a bot handles and which route to an agent and continually evolving your approach as performance data becomes available. It also means watching the metrics that reflect meaningful outcomes rather than a single efficiency number. 

Average handle time is a useful example of why a single number misleads. When AI absorbs the simplest tickets automatically, the interactions left for agents are the harder ones, so average handle time can rise even as the operation improves because human time is going to the customers whose needs actually require it. Performance-managing AI means reading the metrics in context, looking at first contact resolution, customer satisfaction and the quality of the interactions that reach agents, then adjusting the division of work as the business changes.

Frequently Asked Questions

Question: For a specific AI tool, how do you tell whether it is working, and does it matter if it is agent-facing or customer-facing?
Answer: It does matter because a customer-facing tool, such as a self-service bot, is often judged on deflection and whether customers reach resolution without frustration. In this context, the key signals are containment quality and clean handoff to a person when needed. An agent-facing tool, such as real-time assist, is judged on whether it helps the agent resolve the interaction better, so the signals are often resolution quality and agent adoption. In both cases, a single efficiency metric read in isolation can be misleading. The useful read weighs resolution and customer satisfaction against how the tool affects agents and downstream work.

Question: How do you keep an AI customer service tool from giving a confident but wrong answer?
Answer: A frequent cause is fragmented source material, where a tool draws from multiple, sometimes conflicting internal documents and produces something plausible rather than flagging that it is unsure. Part of the answer is governing model access and content centrally so answers draw from current, consistent sources. The other part is automated checking. This means running the system's responses against safety and accuracy checks before they reach a customer, so a wrong or non-compliant answer is caught in testing rather than on a live interaction.

Question: Customer service involves sensitive data. How does TELUS Digital handle the privacy stakes?
Answer: Customer service conversations routinely include financial and personal information, which raises the stakes on how that data is stored, accessed and used to train or fine-tune a model. Privacy obligations differ by industry, geography and what a given company's customers expect, so the work starts with aligning on the specific requirements that govern each engagement. TELUS Digital approaches this through privacy-by-design engineering and enterprise governance. The Fuel iX platform earned the first Privacy by Design certification (ISO 31700-1) for its GenAI-powered customer support chatbot, and it is built to let enterprises manage, monitor and control how generative AI accesses data across the operation. Protecting AI is part of the value chain TELUS Digital operates with clients, built in from the start rather than added at the end.

About TELUS Digital

TELUS Digital crafts unique and enduring experiences for customers and employees and creates future-focused digital transformations that deliver value for our clients. We are the brand behind the brands. Our global team members are both passionate ambassadors of our clients’ products and services and technology experts resolute in our pursuit to elevate their end customer journeys, solve business challenges, mitigate risks, and drive continuous innovation. Our portfolio of end-to-end, integrated capabilities include customer experience management, digital solutions, such as cloud solutions, AI-fueled automation, front-end digital design and consulting services, AI & data solutions, including computer vision, and trust, safety and security services. Fuel iX^TM is TELUS Digital’s proprietary platform and suite of products for clients to manage, monitor and maintain generative AI across the enterprise, offering both standardized AI capabilities and custom application development tools for creating tailored enterprise solutions.
Powered by purpose, TELUS Digital leverages technology, human ingenuity and compassion to serve customers and create inclusive, thriving communities in the regions where we operate around the world. Guided by our Humanity-in-the-Loop principles, we take a responsible approach to the transformational technologies we develop and deploy by proactively considering and addressing the broader impacts of our work. Learn more at: telusdigital.com.