The EU AI Act, which came into effect on August 2, can seem like a legal or compliance issue, best handled by specialist teams or the people developing AI models. But AI is already embedded across the contact center, from customer conversations and agent tools to workforce platforms, quality monitoring, knowledge systems, analytics and workflow automation.
The use of AI in customer service makes AI governance an operating model issue as much as a policy issue.
The EU AI Act takes a risk-based approach, meaning that AI systems are subject to different requirements depending on their intended use and potential impact, and in the contact center not every AI capability carries the same regulatory risk.
An AI tool used to summarize an interaction for an agent, for example, may present a different risk profile from a system that evaluates employee performance, influences customer decisions or operates directly in customer-facing workflows.
Amit Namjoshi, Executive Director of Technology Consulting at TTEC Digital, told CX Today that customer experience leaders have to be involved because customer-facing AI sits directly inside the service experience.
“It’s not a legal compliance and IP issue. There are two paths to a technology platform solution. One is the process element, especially in a contact center. And the other is the technology element.”
The process side is as important as the technology. Legal and compliance teams can define policies, but customer experience leaders need to understand how those policies are applied inside live customer operations.
“Anything that is customer-facing essentially will involve and should involve the CX leaders. And therefore, everything that follows on from there will require that thought leadership from CX,” Namjoshi said, adding simply: “They are the stewards of the AI.”
CX leaders need to understand where AI exists, what each system does, what data it uses and how its outputs affect customers and employees.
Why CX Teams Are Now Part of AI Governance
The EU AI Act is designed to create clearer rules around the use of AI. Once those rules are translated into day-to-day service operations, the practical burden lands with the people running customer experiences.
While a legal team may ask whether a tool is compliant, a CX leader has to answer different questions:
- Where is AI being used?
- What customer data does it access?
- Does it influence a recommendation, escalation, response, or decision?
- Is a human involved at the right moment?
- Can the organization explain what happened if something goes wrong?
- A risk assessment can help CX teams prioritize those questions. AI used for internal productivity or assistance could generally be treated as lower operational risk, while AI that recommends actions or influences customer or employee decisions requires greater attention. Systems that make or influence high-impact decisions, interact directly with customers or handle sensitive data may require the strongest controls and oversight.
The exact classification will depend on the AI system and its intended use under the EU AI Act, so organizations should avoid treating every AI capability in the contact center as having the same regulatory status.
As Namjoshi explained, the Act creates a set of expectations, but operationalizing them requires CX input.
“If you think about the EU AI Act, what is it? It is essentially a bunch of dos and don’ts, policies that have been established to say, ‘this is what we want if you are a compliant organization.’”
The hard part comes next.
“It falls upon the people on the ground to figure out how to implement it. And that squarely lies with the CX leaders to say, ‘how is this going to be governed? How are we going to manage it?’ And set the tone for the process.”
Human Oversight Cannot Be an Afterthought
AI governance in customer experience is about knowing when a human needs to step in.
In some cases, that point may be before a customer-facing decision. In others, it may sit inside workforce management, quality assurance or a supervisor review process. The exact model depends on the use case, but the principle is consistent.
“As a human in the loop, we should have the ability to intervene or take over at the right time. Typically, it would be before the decision point.”
For customer-facing AI, the central question is who approves the output before it becomes part of the customer experience.
“The human oversight is critical at the point of decision-making when it comes to who is essentially pressing the button and saying that this is good to go from an information or a decision perspective.”
Technology Vendors Have a Role to Play
Governance responsibilities also extend beyond the organization deploying AI. Contact center technology vendors have a role in giving customers the information and controls they need to assess and manage their AI deployments.
Vendors can provide documentation about how AI capabilities work, what data they process, where models or services are hosted, how outputs are generated and what controls are available to customers. They can also support governance through configuration options, audit information, testing capabilities, access controls and mechanisms for human oversight.
This makes vendor transparency an increasingly important consideration for buyers when selecting or renewing technology. Procurement teams should be able to understand the AI capabilities included in a platform and the support the vendor provides to help govern those systems effectively.
Compliance, Trust, and Customer Outcomes
The business benefit of AI governance is that risk becomes more visible and easier to explain.
Namjoshi framed the outcome as confidence.
“The risk is never going to go away. It’s always going to be there. It is how well we are able to manage the risk?”
When governance is working, organizations can scale AI with more confidence. They can improve productivity and process efficiency while also improving the accuracy of responses. In a contact center, those improvements should show up in customer satisfaction, quality monitoring and operational performance.
“They will have not only managed to improve productivity and streamline processes, but they will have also seen that responses are more accurate, more importantly, because that will reflect in their CSAT score.”
This is where compliance becomes linked to customer trust. Customers want fast and efficient service, but they also need confidence that the organization is using AI responsibly.
As Namjoshi pointed out: “Scaling of AI broadly is about trust and making sure that it gives you that competitive advantage because everybody’s trying it.”
“The people who are more successful essentially are the ones who have reduced the risk associated with using it extensively and have also gained by using it in the correct fashion,” Namjoshi added.
How AI Regulation Will Shape CX Technology Decisions
Over the next year, AI governance is likely to become more visible in procurement, platform selection and operating model design.
Namjoshi says organizations are already asking different questions in RFPs, workshops and implementation planning.
AI compliance will not sit separately from CX transformation, it will become part of how organizations evaluate platforms, workflows and delivery partners.
Namjoshi said the EU AI Act may help organizations ask the right first question.
“One of the things that will happen as part of the EU AI Act is you will have to explain in your audit where you are using AI.”
TTEC Digital helps clients map AI across the CX estate, identify operational risk and build governance into workflows, platforms, knowledge and agent processes.
Rather than slowing down AI, the aim is to make it safe enough, visible enough and trusted enough to scale. Better governance can give organizations greater visibility into where AI is improving agent productivity, response accuracy, service efficiency and customer outcomes, while highlighting the areas where stronger controls are needed.
As Namjoshi put it:
“For any wise organization, the thing to do is think how about to make it safe, transparent and make sure you’re not infringing anybody’s rights.”
Watch our full interview with Namjoshi, Hidden AI in the Contact Center: What the EU AI Act Means for CX Leaders
