Observations

People are changing their behaviour, expectations, needs and interactions with digital products and services. These changes have consequences for brands, visibility, representation, content distribution, measurement and product responsibility. The analysis follows the observable changes and their effects.

The main phenomena observed, across different levels of GenAI adoption

People are adopting GenAI technologies at extremely uneven levels, depending on where they live, their age and education, and their digital literacy. The observable phenomena can therefore be mapped across different levels of GenAI adoption: this perspective helps us give the right weight to what we observe and weigh its consequences.

1. AI-mediated search

AI Overviews and AI Mode reach billions of people every day inside a familiar path: search engines. Often, there is no intentional adoption of AI. The changes in behaviour at this lighter level of AI are particularly varied:

• Intent-based interaction: from query- and command-based search to search centred on intent
• The customer journey shortens: consideration moves outside the site
• Serendipity declines: targeted results prevent the spontaneous discovery of information
• The conversations you cannot observe: a significant part of the conversations that shape choice happens in private spaces, outside any measurement

The consequences of these phenomena for brands, representation and regulatory obligations are covered separately in Emerging effects.

2. The conversational interface

Daily use of ChatGPT, Gemini and Claude makes chat the implicit new reference point for a substantial share of digital users; the phenomenon is concentrated among younger and more highly educated people. Frequent use of GenAI chat creates new expectations (or intensifies existing ones) toward every digital product:

• The new expectation of zero learning curve
• Chat as the universal default interface
• The user’s language replaces the system’s taxonomy
• From form to progressive refinement
• Memory and context adaptation
• Accessibility and hybrid interfaces

3. Agentic AI: from showing to doing

Some systems no longer only show options: they act on the user’s behalf. The question here is not whether consumer use is already mature (it is not), but how far delegation can go.

• The user expects the system to act, not just to show

4. Cross-cutting themes

Some conditions shape every route: trust, social presence and cognitive load affect how these systems are used and judged.

• Trust, AI literacy and calibration: understanding what AI is and how it works often proceeds in opposing directions, sometimes creating distortions
• Empathy and social presence: an empathetic interlocutor enables a relationship to form
• Cognitive load and wellbeing: response speed and the understanding a user retains are not the same metric

From observations to effects, decisions and roles

Consequences for brands, visibility, distribution, measurement and product responsibility are analysed in Emerging effects. What has to be decided collects organisational choices; What changes for work translates them into operational responsibilities.

What these phenomena rest on

The reflections on these pages rest, as far as possible, on the sources that make up the literature review.