AI is a new tool—you have to learn how to use it

Last updated: 11.11.2026 09:00

The success of AI in customer service also depends on employees’ willingness to engage with it—and on senior management’s ability to prepare for the upcoming shift in roles and make smart decisions about AI governance. The change process is also a communication challenge. Here are a few tips.

The “Contact Center 2026 Trend Study” shows that 96 percent of respondents have a positive view of AI in customer service, but the devil is in the details. A key issue here is not so much a lack of confidence in the technology among top management, but rather concerns about internal resistance. The fear of being replaced by innovative technology leads to defensive reactions on the part of employees. Many CEOs know they need to take action, but they keep putting it off indefinitely. The introduction of AI is therefore first and foremost a leadership and communication challenge. It’s about guiding companies and their employees and showing them that it’s not about replacing people with AI, but about empowering them to work with AI.

Centralized Control Instead of Silos Saves Headaches and Money

Another problem facing many companies is the multitude of isolated solutions that coexist haphazardly. Such siloed solutions generate costs on three levels that reinforce one another.

1. Cost control: Anyone who connects chatbots, voicebots, and email automation individually to different AI models loses track of token consumption and model costs. Often, companies end up paying for three separate LLM integrations, even though centralized control would be more efficient.

2. Compliance risk: The EU AI Act requires traceability and transparency. With siloed solutions, each application must function in compliance with the law individually—meaning data protection, anonymization, and documentation—all of which must be handled multiple times. This is not only expensive but also risky.

3. Switching Costs: Those who rigidly rely on a specific LLM often lock themselves into a technical dependency for years. A middleware approach using an AI gateway solves this problem effortlessly. A unified API enables access to various models, allowing companies to switch between providers such as OpenAI or Google at any time—in a matter of milliseconds. A strategic approach often pays off as early as the second use case.
   Another advantage of the VIER AI Gateway is that it allows companies to define guardrails in advance and specify which model should be used for which scenario. This ensures that complex queries are processed by high-performance models, while standard tasks are handled more cost-effectively. This guarantees not only cost efficiency but also legal certainty and the best possible results for the customer in each case. The gateway thus serves as an intelligent controller that safeguards technological freedom while keeping budgets in check.

Don’t ask what you can do with AI. Ask what AI can do for you!

Anyone looking to get started with AI should never ask first, “What kind of AI do I need?” but rather, “What is taking up most of employees’ time today?” Market surveys show that AI is primarily used to reduce workload. A concrete recommendation, therefore:

1. Companies should identify a clearly defined process—such as standard email inquiries—where the success of AI can be quickly measured.

2. Think in terms of platforms rather than siloed solutions from the very beginning.

3. Involve the workforce from day one and communicate openly. Anyone who tries to work against their own employees will lose. It’s a good idea to ask employees themselves where AI could support them and to gather suggestions.

4. Start small, but with clear success criteria for the first three to six months. And then move forward.

What AI Can Do—and What It Shouldn’t Do

The crucial distinction lies in the question: routine or exception? Today, AI provides reliable services around the clock when it comes to facts, status updates, and scheduling. In companies that use AI, this technology often resolves half of all inquiries completely. However, AI should definitely not be used in situations with genuine emotional implications, complex complaints, or goodwill decisions that involve discretion. Here, humans need to take responsibility. It’s important to note: Customers don’t necessarily always want a human; they want a solution. If a bot provides accurate information at 11 p.m., that’s often more valuable than a conversation with a human the next morning. The smartest approach is a seamless handoff: AI takes over as long as it’s faster, and reliably hands off to a human as soon as the issue crosses a certain threshold.

AI Is Changing the Role of the Service Team

The service team’s role will shift—away from “providing answers” toward “resolving exceptions” and “nurturing relationships.” For a 20-person team, this means that routine tasks are being automated today. In one to two years, the copilot will become a constant companion, suggesting answers and searching for information while the employee remains engaged in the conversation. In three years, AI agents will autonomously handle the majority of standard inquiries. Human roles will shift toward escalation management, quality assurance, and AI training. It is therefore crucial to ask whether and how employees will be prepared for these new roles. Those who want to prepare for this today must communicate transparently, actively involve employees, and make targeted investments in empathy and subject-matter expertise.

EU AI Act – Not a Burden, but an Opportunity

In the short term, the regulatory framework of the EU AI Act is dampening the willingness to invest in AI. Complying with these requirements appears to be a complex undertaking. However, those who rely on European solutions are legally compliant from the outset. And in the long term, regulated AI is actually a tremendous advantage. In five years, the small and medium-sized enterprises that have treated AI as a strategic management decision with a focus on autonomy will have reaped the benefits. Data protection and transparency are key competitive advantages today. Those who prioritize quality and structure now will dominate tomorrow’s market.