AI is no longer competing only for our attention. It is beginning to build relationships with us. What does this mean for companies, brands and leaders?
For more than two decades, digital companies competed for a scarce resource: our attention.
The winning products were those that could attract us, keep us scrolling and bring us back. Search engines, social media platforms, streaming services and mobile applications were designed around one central objective: capture attention and turn it into engagement.
Artificial intelligence is changing the nature of that relationship.
AI does not only want to be noticed. It wants to become useful, familiar and available. It remembers context, adapts its responses and increasingly interacts with us through natural language. In some cases, it is becoming something more than a tool or an interface. It is becoming an interlocutor.
This may mark a shift from an attention economy to an attachment economy.
What is the attachment economy?
The expression is not a formal economic model. It is a useful way to describe a change in the logic of digital products.
The attention economy is built around visibility, frequency and engagement. It asks:
-How can we attract people?
-How can we keep them on the platform?
-How can we make them come back?
A recent study by the French research organisation CRÉDOC illustrates the scale of this shift. According to the study, 23% of people under 40 use AI as psychological support, a friend or a romantic partner. Among those who use an AI friend, 39% say the relationship matters a lot to them. The figure rises to 54% among those who describe the AI as a romantic partner.
These uses do not necessarily replace human relationships. The CRÉDOC study shows that many users of emotional AI also have active social lives. But the technology can still provide a permanent, reassuring and non judgemental space.
This is one of the reasons why attachment is becoming strategically important.
Why are people becoming attached to AI?
Three mechanisms help explain this phenomenon.
- AI feels increasingly human
When a machine uses natural language, remembers previous conversations and responds with apparent empathy, people naturally attribute intentions and emotions to it.
This tendency is known as anthropomorphism. We interpret the behaviour of the machine through human categories because this helps us understand and predict its responses.
- AI creates a space for projection
Users can project their needs, expectations and emotions onto an AI.
The system becomes a space where they can explore ideas, rehearse conversations, ask questions or express feelings without the risks associated with a human relationship.
- AI is always available
AI does not need an appointment. It does not appear tired, distracted or unavailable. It can respond late at night, remember details and adapt its tone to the person in front of it.
For some users, this creates a sense of safety. The AI listens without judging, contradicting or rejecting them.
This is also where the main tensions begin. The CRÉDOC study found that 40% of users of relational AI report feeling always or often lonely, compared with 18% of the rest of the population. Emotional AI may provide comfort, but it does not necessarily resolve the conditions that create loneliness.
AI is changing the relationship, not only the message
Personalisation is not new.
Marketing has used one to one communication for decades. Recommendation engines have made it possible to suggest products based on past behaviour. Customer relationship management systems have helped companies organise their knowledge about individual customers.
Generative AI and AI agents introduce a deeper change.
The system does not only personalise the offer or the message. It personalises the interaction itself.
It can adapt its language, its timing, its level of detail and its recommendations in real time. It can take into account the customer’s history, behaviour, context and immediate needs.
This creates a new kind of relationship between people and organisations. The interface is no longer neutral. Its tone, memory, decisions and boundaries become part of the brand experience.
A customer may not remember which algorithm generated a recommendation. They will remember whether the interaction felt useful, confusing, respectful or manipulative.
What does this mean for business leaders?
The shift from attention to attachment raises four strategic questions.
- What kind of relationship is our AI creating?
An AI assistant can be designed to provide information, complete tasks, recommend products or support decisions.
But it can also create a sense of familiarity and emotional reliance.
Leaders therefore need to define the role of the AI with precision. Is it an assistant, an advisor, a coach, a service representative or a companion? What should it remember? What should it never pretend to understand? When should it hand the conversation back to a human?
These are not only technical questions. They are questions about brand, responsibility and trust.
- How can personalisation create real value?
The next stage of personalisation will be increasingly contextual.
In banking and insurance, for example, signals such as a property purchase, a new child, a change of address or retirement can indicate that a customer’s needs are evolving. AI can help organisations identify these signals earlier and prepare more relevant conversations.
But relevance is not the same as permission.
The value of contextual personalisation will depend on consent, transparency and the ability to explain why a recommendation is being made. If customers feel watched rather than understood, personalisation can quickly become intrusive.
- What happens when trust becomes part of the interface?
AI systems are increasingly involved in conversations that influence decisions.
They may help people choose a financial product, understand a health issue, manage a claim or decide what to do next. In these contexts, trust cannot be reduced to a pleasant tone of voice.
Trust also requires accuracy, explainability, clear limits and access to human support.
The more attached people become to an AI, the greater the responsibility of the organisation behind it. A system that feels like a trusted companion can also become a powerful influence.
- What remains uniquely human?
The rise of AI does not make human relationships less important. It changes where human value is most visible.
AI can inform, educate, simulate and prepare. Human professionals remain essential when situations are complex, sensitive or high stakes. They provide judgment, accountability, empathy and the ability to navigate ambiguity.
The question is not whether humans or AI will own the relationship.
The more relevant question is how humans and AI will share it.
The move from the attention economy to the attachment economy is already underway.
It will affect customer experience, financial services, healthcare, education, leadership and the way organisations build trust.
The companies that understand this shift early will not only design better AI products. They will also make more deliberate choices about the relationships they want to create.
At Innovation Is Everywhere, we design Learning Expeditions and leadership programmes in the United States, Europe and other innovation ecosystems to help leaders understand how AI is changing business and society.
AI is transforming our world. Let’s explore it together.
