The Pressure to Scale
New research from a survey of 200 EU-based contact center leaders finds that AI deployment in customer service is being driven primarily by external pressure rather than internal readiness: 63% cite rising customer expectations, 58% cite competitive pressure, and 56% cite the need to improve operational efficiency. Seven in 10 organizations also said the accuracy and consistency of AI responses is what most influences customer trust, making reliability, not just speed, the actual currency AI deployments are competing on.
Why This Matters for BFSI and Collections
The most significant finding is not about pressure to deploy; it is about who inside the organization believes governance is actually in place. Forty percent of CX leaders said their organization has fully implemented the ability to review or explain AI decisions. Only 24% of compliance leaders said the same. That is not a small measurement gap; it is two functions inside the same company operating on fundamentally different assessments of the same control. For collections operations, where compliance sign-off on explainability is mandatory, this gap is the actual risk. A CX team confident that a voicebot’s decisions are reviewable, while compliance believes otherwise, means the organization does not actually know its own governance state, and that uncertainty surfaces at the worst possible moment: during an audit or a regulatory inquiry.
What the Numbers Do Not Say Out Loud
The survey surfaces a second tension worth sitting with. Eighty-eight percent of respondents said they trust AI more when it supports human employees, yet 71% also said they trust AI when it operates with limited human intervention. Read literally, these two findings pull in opposite directions: AI is trusted more as an assistant, and also trusted more the less a human is involved. That is a signal that trust is being used to mean two different things: comfort with AI as a helper, and comfort with AI as an autonomous operator, without most organizations having clearly decided which one applies to which task. Only about a third of organizations have measures in place to review or explain AI decisions, and just 4 in 10 have imposed controls on what data AI can actually use. The trust numbers are running ahead of the control infrastructure that would justify them.
The Practical Read
For collections and BFSI operations, the CX-compliance gap points to a concrete fix: both functions need to be validating the same explainability standard against the same evidence. That means compliance should be able to pull the same decision logs and audit trail CX points to when it claims a control is fully implemented. Given that nearly 9 in 10 organizations agree AI is more trustworthy with human oversight, the practical next step is making sure CX and compliance are looking at the same record of what the AI actually did before that gap gets discovered by a regulator.