AI in Customer Experience: Why Leadership Judgement Now Matters More Than Technology
By Clare Muscutt, Founder and CEO, Women in CX
AI has quietly crossed a threshold in customer experience.
For several years, it was framed as innovation - something experimental, something exciting, something just over the horizon. That moment has passed. AI is now becoming embedded into the everyday infrastructure of how organisations serve customers and run service operations.
With that shift comes something far more significant than new technology. It brings a new category of leadership decisions.
Inside organisations today, AI isn’t arriving as a tidy strategic roadmap. It’s arriving as a series of decisions landing directly on leaders’ desks. Where should automation sit? What should never be automated? Which vendors should we trust with core operations? And who ultimately carries the consequences if those decisions turn out to be wrong?
These are the conversations now taking place inside CX and contact centre leadership teams around the world.
During a recent WiCX Talk Trends discussion hosted by Women in CX in partnership with Genesys, I had the privilege of exploring this reality with three leaders who are navigating these decisions every day inside global organisations:
Mary Henderson, Head of Contact Center Technology, Reservations and Customer Care at IHG Hotels & Resorts
Christina Bell, Vice President of Contact Center Solutions at BCD Travel
Lorena Lovrić, Director of Customer Experience at Aterian
What followed was not a conversation about tools or hype. It became a much more honest discussion about the weight of the decisions CX leaders are now being asked to make.
When AI Stops Being a Technology Conversation
Much of the public conversation about AI still focuses on capability — what the models can do, how quickly the technology is evolving, and how automation might transform service delivery.
But the reality inside organisations tends to look very different.
AI rarely arrives as a neat capability conversation. Instead, it shows up as a series of trade-offs that leaders must navigate in real time. Decisions about efficiency often collide with questions about customer experience. Automation decisions quickly become employee experience questions. Speed brings operational risk.
As I said at the beginning of the discussion, decisions about AI now shape how customers are served, how frontline teams work, how risk is managed and how trust is built. And they are often being made at pace, with limited precedent and incomplete information.
For CX leaders, this means the challenge is no longer understanding AI. It is exercising judgement about where and how it should be used.
The Pressure CX Leaders Are Facing
One theme surfaced repeatedly throughout the discussion: pressure.
Pressure to move quickly. Pressure to demonstrate efficiency. Pressure to keep pace with competitors who appear to be deploying AI everywhere.
As Christina Bell explained, the challenge is rarely just one of these factors in isolation.
“We definitely feel pressure for speed because we want to stay competitive. We feel pressure for cost because we want to stay sustainable. And we also feel the pressure of how introducing new technology is going to impact our employees.”
These pressures tend to arrive simultaneously. The same initiative that promises efficiency gains may introduce new risks for customer experience. Automation that excites the boardroom can create uncertainty among frontline teams. Leaders are expected to navigate all of these dimensions at once.
That complexity is why leadership judgement has become such an important part of AI decision-making.
“We definitely feel pressure for speed because we want to stay competitive. We feel pressure for cost because we want to stay sustainable. And we also feel the pressure of how introducing new technology is going to impact our employees.”
The Decisions Leaders Are Actually Making
When I asked Mary Henderson what AI decisions were landing on her desk right now, her answer cut through a lot of the noise surrounding AI.
The questions she is being asked to resolve are not theoretical. They are practical.
Where should AI actually sit in the organisation’s ecosystem? Should it be primarily customer-facing, or should the focus be on tools that support agents behind the scenes?
As Mary explained:
“Should we be putting things into voice bots and agentic IVRs, or should we be focusing more on agent tools like real-time guidance, transcription and conversation summaries?”
Alongside that sits a broader strategic challenge. With vendors embedding AI capabilities into almost every platform category, organisations now find themselves navigating overlapping tools and ecosystems.
“The decision isn’t really whether AI is valuable anymore,” Mary said. “It’s where it should live and who we should partner with long-term.”
And then there is the commercial reality every CX leader knows well.
“Contact centres cost,” she added candidly.
But with vendors measuring AI usage in different ways, and capabilities appearing across multiple platforms, even calculating the real cost and value of AI has become more complicated than many organisations expected.
The Efficiency Trap
One of the most revealing moments in the discussion came when Lorena Lovrić described an experience her team had implementing AI-generated email responses during a particularly busy season.
Initially, the results appeared extremely positive. Average handling times dropped significantly, and operational metrics suggested the initiative was a clear success.
But when the team looked more closely, something didn’t feel quite right.
“The responses were technically correct,” Lorena said, “but they felt very transactional.”
In emotionally sensitive situations — complaints or warranty disputes — that tone began to undermine the customer relationship rather than strengthen it.
Instead of scaling the solution immediately, the team paused. They revisited the tone guidelines, strengthened brand guardrails and created clear decision points where agents could modify or reject suggested responses.
The experience highlighted something that many organisations are now beginning to recognise: efficiency is relatively easy to measure. Emotional impact and customer trust are not.
And yet those factors ultimately determine how a brand is perceived.
Starting With the Experience
Another important insight came from Mary Henderson’s approach to evaluating AI opportunities.
Rather than starting with the technology itself, she begins with a much simpler question: what experience are we actually aiming to deliver?
“I start with the experience we’re aiming to deliver, not the technology,” she explained.
When organisations begin with efficiency metrics instead, there is a risk that automation is applied to journeys that were already poorly designed. In those situations, automation simply amplifies the friction that customers were already experiencing.
As Mary pointed out, pushing customers through digital channels may improve containment metrics, but if those journeys require customers to repeat information or navigate unnecessary steps, the experience has ultimately deteriorated.
Automation does not fix poor design. In many cases, it magnifies it.
“I start with the experience we’re aiming to deliver, not the technology”
Why Agents Still Matter
Across the discussion, one principle remained consistent: AI should support frontline teams, not replace them.
For Lorena Lovrić, that philosophy is captured in a simple phrase she often uses with her teams.
“Happy agents equal happy customers.”
Her leadership approach is unapologetically agent-first. Every AI decision, she believes, should be evaluated through the lens of whether it improves the experience of doing the job.
If a tool increases cognitive load, creates additional complexity or reduces agent confidence, adoption will struggle regardless of the technology’s theoretical benefits.
For this reason, many organisations are starting with agent-assist capabilities first — tools such as transcription, conversation summaries and real-time guidance that reduce friction in the agent experience without introducing additional risks for customers.
The Courage to Slow Down
One of the most important leadership themes that emerged from the discussion was the courage required to say “not yet”.
Mary Henderson described situations where new AI capabilities were technically available but operationally premature.
“We could turn those features on tomorrow,” she said. “But we’re not operationally ready.”
Governance structures may not yet exist. Agents may not have been trained. Operational teams may not fully understand how the technology will affect day-to-day workflows.
Deploying technology too early can create challenges that are far more difficult to resolve later.
And when systems fail in customer operations, the consequences are usually felt most directly by frontline teams.
As Mary reflected:
“When tools work well, it’s joy for agents. When they fail, it’s cruel.”
The Real Differentiator
Perhaps the most striking reflection came from Lorena Lovrić towards the end of the discussion.
“AI is no longer the differentiator,” she said. “Everyone has access to the same models. The real differentiator is judgement.”
Technology will continue to evolve. New capabilities will continue to appear.
What will ultimately distinguish organisations — and the leaders guiding them — is the judgement used to decide where AI should be applied, and where it should not.
The future of customer experience will not be determined simply by how quickly organisations deploy AI.
It will be shaped by how thoughtfully leaders decide what those systems are allowed to optimise, and what they must protect.
Customers. Employees. Trust.
Those are not technical considerations. They are leadership ones.
And they are now among the defining leadership challenges of our time.
Stay Connected with Women in CX
This conversation was part of our WiCX Talk Trends series, where we bring together leaders from across the industry to explore the real challenges shaping customer experience today.
A huge thank you to Genesys for supporting this conversation and helping ensure women’s voices remain central to these industry debates.
We host free webinars and discussions like this throughout the year for women and allies working in CX and technology.
If you’d like to stay up to date with upcoming events and community insights, you can join the Women in CX community.