Front pagesponsored · Diabolocom
CX AI21m · 10:31 BST · 4 min read
Is Contact Center AI for Augmentation Missing the Opportunity?
89% of contact center leaders expect humans to remain central to customer support. So why are employee training and workforce readiness consistently ranked as the lowest priorities in AI investment decisions? The gap between stated intention and operational reality is striking.
Most contact center leaders will tell you that AI should be used to augment agents, not replace them.
In an ideal world, the agent of the future will handle complex conversations, build relationships, and exercise the kind of judgment no virtual agent can replicate.
The belief among leaders that this is the best use of AI seems genuine.
Indeed, a new report from Diabolocom and CCW Digital, based on interviews with 100 US CX leaders, finds that 89% expect humans to remain central to the support process. In contrast, just 11% believe AI agents will meaningfully replace human agents on the service frontline.
However, the same study reveals an industry that is investing in AI technology faster than it is preparing the people those tools are supposed to empower.
The gap between stated intention and operational reality is striking.
The Environment is the Problem
It would be convenient to frame this as a technology problem. If you buy the wrong tool and deploy it badly, then naturally, your agents will struggle.
But the report points to something more structural. 82% of leaders say their employees access multiple systems throughout a single workday.
74% require agents to switch between screens within a single live interaction, while only 24% of contact centers run on a fully integrated platform.
These are not AI failures; they are the consequence of introducing AI before fixing the underlying processes and systems.
Indeed, Frédéric Durand, CEO ofDiabolocom, made this point in his recent manifesto:
“The breakdown appears in how the industry packages, positions, and deploys technology into live operations. The gap does not stem from a lack of AI power. It comes from a lack of discipline in how we apply it.”
The consequences show up in the workforce readiness numbers. Despite two years of sustained AI investment and near-universal confidence in its potential, only 17% of leaders believe their employees are ready to handle what they describe as next-generation work.
That figure is almost identical to a 2024 CCW Digital finding, which means organizations have largely stood still on this while the technology has raced ahead.
Who is Making the Decisions
Part of the explanation for this breakdown lies in who is actually driving AI strategy.
61% of final automation decisions are made wholly outside the contact center function. The people who understand the operational reality – the agents, supervisors, and team managers dealing with fragmented systems and rising call complexity – are not at the table when the technology gets chosen.
Speaking in the report, Chris Dorn, Chief of Staff at Bison Wealth, said:
“You’ve got to be careful, because a lot of people use the term AI and don’t really know what they’re saying.”
The result, predictably, is that employee experience gets deprioritized.
The report reveals that only 34% of organizations measure employee sentiment about AI’s impact on their workflows.
Moreover, employee training and time-to-proficiency are the two lowest-ranked considerations in AI investment decisions. At the same time, 77% of organizations have not seriously re-evaluated compensation or career paths in light of what they are now asking their agents to do.
What the Successful Deployments Look Like
Despite the facts and stats discussed thus far, the report is not completely bereft of any optimism.
The organizations getting their AI strategy right share a consistent approach:
Culligan is a clear example of this strategy playing out in real terms.
Rather than automating the agent’s entire role, the company identified the most painful task (voicemail processing) and targeted that. Generative AI now transcribes each message, assesses urgency, and drafts the ticket.
The result was a 70% reduction in voicemail processing time, an 85% first-contact resolution rate, and an NPS score described as an industry record.
Nikon Optical Canada took a similar approach. Before deploying AI, they integrated their contact center with Salesforce so that every call automatically pulled the customer record. Response time dropped from 60 seconds to 20 seconds.
A subsequent analysis found that 50% of all calls were a single automatable task, giving them a clear, data-backed starting point for what to target next.
The report’s recommendations follow the same logic:
-
Address the governance and workforce implications of AI transformation before they become a problem, rather than after
The contact centers closing the gap between AI ambition and operational reality are not deploying more AI; they are being considerably more deliberate about where, why, and for whom they deploy it at all.
Read nextordered by techtelligence · every pick explained
more from Diabolocom · sponsored
FCC Offshore Call Center Proposal: Why Smart Leaders Are Already Preparing
28 Jul 2026same beat · CX AI3 Consumer Expectations Defining AI in CX This Week2hsame beat · CX AIHow CX Leaders Should Own AI Agent Performance21h
More from the CX AI desk
News
3 Consumer Expectations Defining AI in CX This Week
AI is raising customer expectations faster than many CX teams can meet them. New research from Zingtree, NTT DATA and Alchemer shows consumers want speed and convenience, but still expect control, context, and a clear path to human support for complex moments.
Sean Nolan · 2hSponsored · Centrical
How CX Leaders Should Own AI Agent Performance
CX leaders must close the orchestration gap between AI agent activity and measurable customer and business outcomes. Centrical CEO Gal Rimon explains how hybrid human-AI teams, governance, reskilling, and real-time optimization can drive stronger execution
Rob Wilkinson · 21hSponsored · Parloa
Why Customer Service AI Needs a Post-Launch Operating Model
Parloa’s Csaba Tamas tells CX Today that launching an AI agent is only stage one. Business teams need practical control after go-live, treating agents like employees who need coaching, feedback and constant fine-tuning to stay aligned with the business
Nicole Willing · 9 Oct 2026
