At MST, we have spent three decades in this industry. We have seen fax, CRM, live chat, WhatsApp Business… and with every wave, someone said, “This time, it really will change everything.” With artificial intelligence, for the first time, we have to agree. But not for the reasons most people think.
AI is not replacing the contact center. It is forcing it to rethink itself from the ground up. And for those of us working in customer service, that is far more interesting news than “robots are going to take our jobs.”
What the numbers tell us (and what they don’t)
Let’s start with the data, because there is plenty of talk about AI in this industry, but far less scrutiny.
In Spain, the contact center industry generated €2.025 billion in revenue, up 0.94% on the previous year, according to the CEX Association’s 2025 Market Study. That figure is interesting in itself, but another catches our attention even more: technology investment as a share of company costs rose by seven percentage points to 68%. Companies are not talking about AI simply because it is fashionable; they are committing real budgets to it.
And where is that investment going? According to the same study, robotic process automation (RPA) and artificial intelligence have already been implemented in 89% of CEX member companies. That is a remarkably high figure, and it confirms something we repeat in every meeting with our clients at MST: AI is no longer a pilot project tucked away in an innovation department. It is core infrastructure for the modern contact center.
At the national level, a 2026 AWS study indicates that 61% of Spanish companies have already implemented an AI solution, eleven percentage points more than the previous year. According to the report, Spain is seven percentage points above the European average for adoption.
So far, it all sounds very promising. And this is where, with 30 years of experience behind us, we need to hit the brakes for a moment.
The figure hardly anyone mentions: investing is not the same as getting results
A recent study reported by MuyCanal, examining more than $250 billion in global AI investment in 2025, reveals that only a quarter of companies say they have achieved a significant business impact from that investment. It bears repeating: three out of four companies invested in AI without seeing a meaningful impact.
Why does this happen? Because many companies treat AI as a patch added to existing processes, rather than redesigning those processes from scratch around how AI and people will work together. In a contact center, that distinction makes all the difference. There is no point putting a chatbot in front of a poorly designed process: all you achieve is automating customer frustration, not eliminating it.
This is exactly what we keep saying at MST: technology does not fix a bad process; it amplifies it. Before automating, you need to truly understand the customer journey.
From trend to strategic decision
What is genuinely changing in 2026 is the role of conversational AI within organizations. It is no longer solely the responsibility of the innovation department: it has become a leadership decision, with a direct impact on customer experience, operations and business results, according to PwC’s 2025 Contact Center Trends Study, published by FUNDAE.
The difference between companies that benefit from AI and those that do not is no longer whether they have a bot. It is whether that technology is part of a coherent conversational architecture: one where voice, WhatsApp, web chat and social media stop operating in silos and start sharing customer history and context in real time.
For us, that is the real revolution. It is not that a machine can speak like a person. It is that customers finally no longer have to repeat their problem three times across three different channels.
Will AI replace human agents?
This is the question we are asked most often, and the short answer is: not in the way most people fear.
What is happening is a real redesign of the agent’s role. AI is increasingly handling repetitive, low-value tasks first: balance inquiries, appointment changes and frequently asked questions. This frees human agents to focus on what truly requires judgment, empathy and decision-making: complex interactions, sensitive complaints and customers who need to feel truly heard.
Gartner predicts that by 2029, agentic AI could autonomously resolve up to 80% of routine customer service inquiries. This deserves to be stated clearly: it is a forecast, not an established fact, and as with any prediction looking several years ahead, we treat it with some caution.
What is already clear today is that the model delivering results is a hybrid one: AI for volume and repetition, people for judgment and emotional connection. Neither entirely human nor fully automated.
How to avoid becoming part of the 75% that sees no results
With all of this in mind, at MST we sum up our recommendations in four principles:
- Start with the process, not the tool. Map the actual customer journey before deciding what to automate.
- Measure from day one. Track first-contact resolution, waiting times and customer satisfaction, not just cost savings.
- Train your people for their new role. Agents handling complex complaints today need different tools and training from those they needed five years ago.
- Do not buy technology out of fear of falling behind. According to several recent studies, technology FOMO is one of the main reasons AI investment fails to generate a return.
A final thought
Thirty years in this industry have taught us one thing: technology trends come and go, but what endures is the quality of the conversation between a company and its customers. AI does not change that fundamental truth. What it does is force us to decide, far more clearly than before, which parts of that conversation are worth automating and which unquestionably need to remain human.
*Note: All figures are verifiable as of this article’s publication date (September 2026).