Humanising AI — how to train your AI voice for authentic customer service

After more than 20 years of training and coaching Customer Service Professionals, I received a first-time request: a client asked me to "train" their AI chatbot using the same framework I use with their human team. The goal? Bring consistency, professionalism, and warmth to every customer interaction—whether it's human or AI-powered.

The timing couldn't have been better. As more brands embrace AI in customer experience, the real challenge isn't what AI can automate—it's how AI communicates. Customers are willing to interact with AI, but only if it feels respectful, clear, and honest. That's what I call humanising AI.

The 3-F Formula — my foundation for AI communication

Over the years, I developed an easy-to-follow framework for customer service communication: the 3-F Formula.

Friendly tone + Formal words + Focused conversation

It’s the same formula I use to train Customer Service Professionals—and now, it’s being applied to AI systems as well. When humanising AI, this often means changing words like “thanks” to “thank you,” replacing “no problem” with “you’re welcome,” and avoiding overly casual phrasing. The result is communication that feels customer-centered and consistent with your brand’s voice.

This approach aligns perfectly with what I see in Gladly's AI capabilities. Their Sidekick feature is designed to support authentic, brand-aligned interactions—ensuring customers never feel like they're speaking to a disconnected, robotic system.

Three questions to guide Your AI strategy in CX

Once you've applied the 3-F Formula, it's time to go deeper. Here are three strategic questions I recommend when building or refining your AI-powered customer service experience.

1. Will you disclose that it's AI?

According to Gladly's recent consumer research, most customers prefer clear disclosure when they're interacting with AI. Yet many brands still allow their AI to say "I" and "we," creating a false impression of human interaction.

Here's what I recommend instead: be transparent. For example:

"Please enter your phone number in case one of our Specialists needs to contact you."

No deception needed—just clear, respectful communication. AI transparency builds trust, not frustration.

2. Should AI attempt to express empathy?

Empathy tops every customer service skills list. But here's the thing: AI can't truly feel empathy, and customers know it.

Some brands train their AI to use phrases like "I understand how frustrating that must be," but this can backfire. When humanising AI, authenticity matters more than mimicking human emotions. Let AI handle transactions and logistics efficiently, and leave emotional nuance to the Customer Service Professionals who are trained to recognize and respond to real human feelings.

3. Who owns ongoing AI training?

Training your AI voice isn’t a one-time project—it’s an ongoing process, just like training for your customer service team. As processes evolve, new products launch, or customer feedback rolls in, someone needs to ask:

"Should this be handled by AI, AI + human, or all human?"

I recommend calibrating AI interactions alongside your regular call reviews. Make sure the language your AI uses aligns with both your brand voice and what customers actually want to hear.

Why authentic AI matters for the future of customer experience

Retailers are integrating AI in customer service at record speed. But efficiency without authenticity leads to frustration. What customers want are genuine interactions across every touchpoint—not AI pretending to be human.

This is exactly why I appreciate Gladly's approach to Customer AI. Their system maintains memory across channels and interactions, ensuring that whether customers engage with AI or a member of the service team, it feels like one continuous, authentic conversation.

My final recommendation — ask your customers

Whether you're refining your AI chatbot, deploying a new platform, or rethinking your entire approach to AI in customer service, one principle should guide every decision:

Ask your customers.

Ask them what they want from AI. Ask how current interactions feel. Ask what would make their experience more meaningful. Their answers are your best blueprint for humanising AI—and for building customer service that's not just efficient, but genuinely helpful.

After all, the future of AI in customer service isn't about replacing human connection—it's about making every interaction, whether human or AI-powered, feel authentically yours.

About Charlotte Purvis

Charlotte Purvis has spent 25+ years analyzing nearly every word her clients use when speaking with customers. She developed the 3-F Formula to give them a systematic approach to customer communication. The formula is featured in her book, The Customer Communication Formula: How to Communicate with Your Customers and Boost Your Customer Service Brand.

Charlotte has recently created an online course based on her book. She will “gladly” join your podcast, webinar, or company event for a conversation about customer communication.