As AI tools become increasingly integrated into customer service operations, contact centres must equip their agents with the knowledge and skills to use them effectively.
However, many companies are still playing catch-up, rapidly deploying AI solutions without providing sufficient training.
The key to success lies in treating AI as a collaborative partner rather than just another technology roll-out. Agents should feel empowered and confident using AI, understanding how it enhances their role rather than replaces it.
To find out more, we spoke to Brad Cleveland to discover how well contact centres are currently handling AI training, what topics should be covered, how often training should be updated, and how to measure its effectiveness.
Video: How to Train Agents to Work Alongside AI
Watch the video below to hear Brad explain how contact centres can successfully train agents to work alongside AI:
With thanks to Brad Cleveland, a well-respected consultant, keynote speaker, and course instructor, for contributing to this video.
The Current State of AI Training in Contact Centres
While AI adoption is accelerating, many contact centres are not investing enough in training agents to use these tools effectively.
Instead, they focus on rolling out AI systems, expecting agents to adapt on their own. This leaves employees feeling unprepared or even resistant to the technology, rather than seeing it as a tool that enhances their work, as Brad explains:
“Most organizations are playing some catch-up. While they’re rapidly deploying AI tools, in many cases they’re often not investing enough in helping agents really understand how to work effectively with these new tools and truly feel empowered by them.
You know we want them to be excited about them. Think of a good example, maybe a restaurant upgrading from traditional kitchenware to one with smart cooking equipment.
You wouldn’t just install the equipment and expect the chefs to figure it out. They need to understand how the technology works, and how to blend their expertise with it.
So, the organizations that are succeeding are those that treat AI tools as a partnership between their employees and the technology, not just another technology deployment.”
Simply installing the technology isn’t enough. Contact centres that treat AI as a partnership between technology and employees, rather than just another tool, are the ones seeing the most success.
Three Essential Topics for AI Training
For AI training to be truly effective, it must cover three key areas that help agents not just understand AI, but also use it confidently in their daily tasks:
- Understanding How AI Works
- Practical Skills for Working With AI
- Critical Thinking and Decision-Making
By covering these three areas, contact centres can ensure that their agents aren’t just using AI passively but are actively leveraging AI to enhance their expertise and improve customer interactions.
1. Understanding How AI Works
Agents don’t need deep technical expertise, but they should grasp how AI processes information, detects patterns, and generates responses.
When they understand the logic behind AI decision-making, they can use it more strategically and trust its outputs.
2. Practical Skills for Working With AI
A crucial part of AI training is helping agents know when to rely on AI and when to take control themselves. One of the most valuable skills in this area is prompt engineering, the ability to ask AI the right questions to get the most effective responses.
“The practical skills of working alongside AI, when to rely on AI, when to take the lead, what prompts get the best responses.
The AI world came up with the term ‘prompt engineering’, which sounds like something IT or management does, but it’s something we all need. Prompt engineering, how do we ask the right questions so that we get the right responses from AI.”
Since AI outputs are only as good as the inputs it receives, agents need to be trained on how to craft and refine their queries to get the best results.
3. Critical Thinking and Decision-Making
AI isn’t infallible, and agents need to be able to identify when AI-generated responses may not be appropriate or fully accurate.
For example, an AI tool might provide a technically correct solution, but if a customer is frustrated or emotional, the agent must recognize when a more empathetic, human-centred approach is needed, as Brad explains:
“The critical thinking skills to spot when AI might be off track, or where human input would create just the right experience for the customer.
An example could be AI is suggesting resolution steps for a technical issue, and that’s great and they’re accurate, let’s say, but we need to know when to go off-script for a customer that’s really stressed, there’s some emotional distress, or some context that the tool is not going to understand and respond to appropriately.”
Training should emphasize how to balance AI efficiency with human judgement to ensure the best customer experience.
How Often Should AI Training Be Updated?
With AI tools evolving rapidly, traditional annual or biannual training cycles are no longer sufficient.
Instead, contact centres should adopt a continuous learning approach, with formal updates at least quarterly – or even more frequently if the AI system undergoes significant changes, as Brad explains:
“I recommend a continuous learning approach with formal updates at least quarterly. That would be ideal, maybe more often than that if a tool’s developing very quickly, but at least quarterly.
It’s a bit like keeping up with smartphone updates, you can’t just learn a smartphone once and you’re done. Every new feature or capability that we really need to use, we need to understand and know how to integrate into our daily work.
Many of these training cycles and efforts will be quick and easy, and it gets easier over time, but the process needs to be ongoing.”
Agents need regular opportunities to stay up to date, and some contact centres are successfully implementing collaborative learning models to make this process easier.
By integrating training into everyday workflows, contact centres can ensure that learning is ongoing, relevant, and immediately applicable.
“I’m also seeing some organizations create opportunities for agents to share what they’re learning, discuss challenges and learn from each other.
One contact centre has an AI Insight session every Friday where agents share best practices and aha moments from the week, so that helps create some camaraderie and some real learning in that.
It’s really about bringing that learning to life and making sure that it’s continuous.”
How Do You Ensure AI Training Is Effective?
To ensure AI training is truly effective, contact centres need to look beyond traditional customer service metrics.
“The key for most organizations is to quickly move beyond any traditional metrics. I mean track the importance of quality and customer satisfaction, those are important traditional KPIs, but also look for signs that agents are using the tool strategically, not just mechanically. Like sports training, you need both practice drills and real game experience.”
While KPIs like customer satisfaction and response time are still important, a well-trained workforce should also demonstrate that they are using AI strategically, not just mechanically.
By moving beyond a “one-and-done” training mindset and focusing on strategic, continuous, and hands-on AI training, contact centres can ensure their agents are not just using AI tools, but are actively leveraging them to enhance their roles and create the best possible customer experience.
Key indicators of training effectiveness include:
Coaching and Observation
One of the best ways to measure AI training effectiveness is through regular coaching sessions, where supervisors observe how agents interact with AI tools in real customer interactions.
However, this should be a two-way process – agents often discover new best practices that even their supervisors might not have encountered yet.
Encouraging open discussions between employees and management fosters innovation and keeps training relevant.
“The most effective programmes include regular coaching sessions where supervisors will observe how agents are interacting with these tools in real situations and provide helpful coaching.
And that’s two-way coaching. I mean our employees are using these tools every day, they can quickly learn things that their immediate supervisors may not have even used yet. So that’s got to be a two-way conversation.”
AI–Human Harmony Moments
A key indicator of success is when agents take AI-generated responses and personalize them to enhance customer interactions.
This demonstrates that they are using AI thoughtfully rather than relying on it passively.
Encouraging agents to articulate their decision-making process also helps ensure they are confident in knowing when to use AI and when to step in with their own expertise, as Brad explains:
“The idea is we want everyone to use the best practices that are emerging. Look for what some call AI–human harmony moments, when an agent takes an AI-generated response and personalizes it with what we do as humans best. Those are moments that show training is really working.
And see that agents can explain their decision-making process, and that they’re confident in knowing when to rely on AI, and when to really take the lead. These are all indicators of effective training.”
Practical Application and Continuous Improvement
Just like in sports, where both drills and real gameplay are essential for skill development, AI training should blend structured learning with hands-on experience.
Contact centres should also create feedback loops, where training is refined based on agent insights and real-world results. This approach ensures that AI training remains relevant, effective, and continuously evolving.
If you are looking for more great insights from the experts, check out these next:
- Tips to Tackle Agent Burnout in Contact Centres
- Agent Self-Evaluation – The Pros and Cons
- The Secrets to Writing Better Customer Emails
- What’s the Best Sample Size for Quality Scoring?
Author: Robyn Coppell
Reviewed by: Xander Freeman
Published On: 22nd Apr 2025 - Last modified: 23rd Apr 2025
Read more about - Skills, Artificial Intelligence, Brad Cleveland, Top Story, Training, Videos
