Artificial intelligence has become the big talking point of the moment – a trend that has now lasted four years and even featured in a 2024 New Year’s Eve special on RTVE, which humorously parodied the rise of AI. There is talk of models, assistants, co-pilots, agents, automation, productivity and new use cases. But as the technology matures, a less obvious yet increasingly important question is beginning to emerge: how to turn all that potential into services that are understandable, customisable and meaningful for customers, businesses and users.
Because it is one thing to have AI capability, and quite another to transform it into a useful, understandable, secure and scalable solution.
AI will not reach the market solely through the power of its models or the speed of technological innovation. It will arrive when it is integrated into specific experiences, with a clear business model, a simple user experience, privacy guarantees, support, measurement and trust.
From technological promise to real-world service
For years, many digital innovations have followed a similar pattern. First, a technology with enormous potential emerges. Then applications, providers, devices, services and business models emerge. And, over time, the market begins to organise that complexity around simpler offerings for the user.
The smartphone is a good example. When it first appeared, it wasn’t clear just how much it would ultimately enable beyond the device itself. First came the technology: a touchscreen, mobile connectivity and a new way of interacting with the internet. Then a vast ecosystem of apps, developers, accessories, services and digital experiences emerged. And, over time, that complexity became something very straightforward for the user: from opening an app store to finding a solution for communicating, shopping, getting around a city, learning, playing, making payments or managing their health.
With AI, we are entering a similar phase. The challenge is no longer simply to identify what artificial intelligence can do. The challenge is to decide where it truly adds value, for what type of customer, within what kind of experience, with what level of responsibility and under what business model.
It is not enough to say that a product ‘has AI’ – almost everyone boasts about that now. It has to better address a need. It can help an SME automate tasks, streamline customer service, better protect a family, enhance an educational experience, optimise a professional’s productivity, personalise content or make an internal process more efficient. The difference between an interesting feature and a service with real potential will lie in the clarity of the benefit and the ability to bring it to market consistently and sustainably.
AI is already beginning to function as a new layer of services
Artificial intelligence is moving beyond being merely an isolated feature to become a cross-cutting layer of digital life. It can be integrated into devices, communications, productivity, customer service, entertainment, security, education, the smart home or business services.
This changes the business logic. Until now, many digital offerings have been packaged as subscriptions, benefits included in a deal, or additional services. With AI comes a new level of complexity: it is not just access that matters, but also usage, computing costs, response quality, privacy, traceability and the ability to integrate into a real workflow or everyday life.
Furthermore, AI assistants and agents are beginning to capture an increasing share of users’ time. Previously, many tasks were carried out within apps, search engines, marketplaces or platforms. Now they are increasingly being handled via conversational interfaces or more automated experiences. This raises a strategic question for many companies: how to remain relevant as the customer interface shifts towards AI.
From content bundles to capability hubs
Telecoms companies are well acquainted with the logic of digital bundling. Over the years, they have moved from selling connectivity to integrating entertainment, security, devices, cloud, gaming and business solutions. In many cases, their value has not lain in creating each service from scratch, but in making it accessible, packaging it, integrating it into channels, simplifying payment, providing support and adapting it to the local market.
With AI, this role may become even more significant. The opportunity lies not simply in reselling a specific tool, but in building a portfolio of capabilities: assistants, automation, productivity, security, analytics, content generation, intelligent customer service or support for small businesses.
The question, therefore, is not just which AI service is incorporated into an offering. The question is what need is being addressed, what experience is being designed, how usage is measured, how costs are controlled, what guarantees are offered, and how all of this is integrated into the customer relationship.
The new challenge: measuring, governing and explaining usage
AI introduces a new economic unit distinct from that of other digital services. In connectivity, the user understands concepts such as data, speed, coverage or devices. In video, they understand access to content. In cybersecurity, they understand protection. In AI, value may depend on a specific task, the volume of usage, the type of query, the capacity for automation or the level of personalisation.
Therefore, turning AI into a service requires new measurement and governance capabilities. It is essential to understand how much it is used, how much it costs to deliver each type of interaction, what value the customer perceives, the level of repeat usage, and which business model ensures the offering is sustainable.
It also requires very careful design in terms of privacy, consent and trust. AI works with context, data, instructions and results. If it is to be scaled up, it is not enough for it to simply work technically: it must be explainable to the user, governable by the company and consistent with regulatory and reputational expectations.
The distinctive role of a telco
Telcos occupy a particularly interesting position in this context. They are often analysed solely in terms of infrastructure, but their potential role in the AI economy may be broader.
A telco provides connectivity, scale, commercial reach, physical and digital channels, billing capacity, customer service expertise, regulatory knowledge, a local presence and an ongoing relationship with millions of customers. But, above all, it offers something highly valuable when a technology begins to become widespread: the ability to turn a complex trend into an understandable commercial proposition.
For many technology players, reaching the end customer is no easy task. They may have unique capabilities, but they need distribution, local adaptation, credibility, commercial support and the ability to scale. For a telecoms company, collaborating with the technology ecosystem allows it to accelerate entry into new categories without having to build everything in-house.
The opportunity lies not in replacing AI specialists, but in connecting capabilities with real market needs. And that requires more than simply identifying good technologies. It requires understanding where demand lies, which segments might value the offering, which operations wish to drive it forward, and what conditions make its launch viable.
It is not enough simply to add capabilities: they must be orchestrated
In an ecosystem as fragmented as that of AI, it can be tempting to think that more capabilities mean more value. But this is not always the case. The value lies not in accumulating options, but in building a clear service logic.
Orchestration means selecting carefully, prioritising judiciously and being clear about the role each component plays within the value proposition. It also means deciding which capabilities should be in-house, which can be accelerated through third parties, and where it is necessary to maintain control over the customer relationship.
And, sometimes, it means letting go. Not all use cases merit the same effort. Not all integrations deliver differentiation.
Not all propositions justify the complexity they generate.
In AI, this discipline will be particularly important. There will be general-purpose models, vertical solutions, productivity tools, device-embedded services, cloud capabilities, process automation, advanced analytics and conversational experiences. The advantage will not lie in working with everything, but in building combinations that make sense for the customer, for operations and for the business.
From partnership to impact
The success of an AI proposal should not be measured solely by its launch onto the market. It should be measured by its actual impact.
Does the customer understand the proposal? Do they use it? Do they perceive value? Do they trust the service? Does it improve the experience? Does it reduce friction? Does it generate revenue, loyalty, efficiency or differentiation? Can it be scaled to other segments or markets? Is the operating model sustainable?
These questions are less eye-catching than the announcement of a new partnership, but they are what determine whether an initiative works. In digital services, the details matter a great deal: the way customers sign up, activate the service, pay, cancel, receive support or resolve an issue can completely change their perception.
In AI, moreover, particularly sensitive issues come into play: privacy, security, explainability, accountability and trust. That is why bringing AI to the customer is not just a matter of innovation. It is a matter of service design, commercial execution and governance.
A new business capability
Artificial intelligence is accelerating a transformation that was already underway: companies no longer compete solely on what they are capable of building in-house, but on what they are capable of connecting, integrating and scaling.
This does not mean giving up in-house capabilities. On the contrary. The clearer a company’s unique assets are, the better it can decide where to build, where to collaborate and where to retain control.
For a telecoms company, the strategic question is not merely which AI to incorporate. The question is what role it wishes to play in this new digital and industrial wave: whether to limit itself to providing connectivity for services created by others, or to aspire to become a trusted platform capable of delivering AI solutions to customers with scale, context and proximity.
The difference between these two positions will not depend solely on technology. It will depend on the ability to identify relevant trends, validate demand, design a clear commercial proposition, integrate it into channels and execute it rigorously.
AI could be one of the major forces of transformation in the coming years. But for it to truly reach people and businesses, it will take more than just powerful models. It will require integration, trust, measurement, execution and a clear vision of how to turn technology into a service.
And therein may lie one of the competitive advantages for telcos. In an increasingly complex market, true value will not lie solely in developing new technological capabilities, but in making them accessible, useful and relevant to customers and organisations. This approach aligns with Telefónica’s ambition to“become the best gateway for citizens to digital technologies”, by offering innovative solutions that simplify technological complexity and generate tangible value. Because, ultimately, the success of any innovation will not be measured solely by its technological sophistication, but by its ability to improve the customer experience and build trust. As the company’s value proposition states, “the quality of our services will be measured by the trust of our customers”
