The Hyper-Targeted Attention Economy: Paid to Watch Ads Generated for Us

For years we have given away our data in exchange for free services. Today, with Artificial Intelligence capable of generating video ads tailored to the individ

For twenty years, the founding pact of the Internet has remained unchanged and silent: the user receives access to free global services (search engines, social networks, maps) and in exchange gives up their data and their attention to advertisers. It is an indirect exchange, in which we never see a penny change hands. Today, the intersection of Generative Artificial Intelligence and the attention economy is pushing the market toward a radical provocation: what if platforms started explicitly paying us monetary compensation to watch ads generated tailor-made, exclusively for us?

The idea of directly remunerating the user for their time may seem, at first glance, a democratization of the web. Finally we would be paid for the value we generate. Yet, in this in-depth piece for the Scenarios and Reflections column, we will dismantle this utopia. Analyzing the frontiers of generative marketing and the recent vetoes of European privacy regulators, we will explore why compensation in micro-transactions does not automatically transform hyper-personalized advertising into a fair exchange.

Attention is not just a temporal metric. It is the bulwark of our decision-making autonomy.

1. The Generative Incentive: The Effectiveness of "Audience of One" Advertising

Why should platforms push themselves to generate individual content (and perhaps pay us to watch it) instead of broadcasting the same spot to millions of people? The answer is in the numbers.

A recent field experiment, published in the journal Marketing Science ("Generative AI and Personalized Video Advertisements"), measured the impact of GenAI on advertising [1576]. The study finds that personalized video ads dynamically generated via AI increase engagement (the interaction rate) by as much as 6–9 percentage points compared to generic video formats or simple personalized images.

Artificial Intelligence has slashed production costs: today it is economically sustainable to generate a million different video spots for a million users, modulating tone of voice, colors, and argumentative levers to each person's exact psychological inclinations. The commercial incentive to create an audience of one (an audience made up of a single person) is simply too profitable to be ignored.

2. The Paradox of Relevance: "Creepy, Not Useful"

The fact that a tailor-made ad is statistically more effective does not mean it is desired. The "US Personalized Advertising Report 2025" conducted by YouGov offers a very harsh empirical counterpoint: 54% of US respondents say that personalized ads are "creepy" (creepy) [1582].

This figure reveals a fundamental misunderstanding of modern marketing: "more relevant" does not automatically equal "more acceptable".

A large study published in Springer ("Can I have it non-personalised?") conducted empirical tests on over 3,400 consumers, revealing a deep reluctance to share data, even when superior services are promised in return [1574]. The user's willingness depends on the very delicate balance between the perceived value of personalization and the perceived psychological cost of losing privacy (the so-called privacy-personalisation trade-off). If, in order to see an ad that interests me, I have to know that an algorithm has analyzed my private messages or my heartbeats, refusal prevails over utility.

3. The Price of Consent: How Much Is Our Autonomy Worth?

Faced with this reluctance, the tech industry could play its ace: direct remuneration. But what monetary compensation would we consider fair for giving up identifiable data?

A study published in Decision Support Systems investigated consumers' so-called willingness-to-accept (willingness to accept compensation) in exchange for giving up privacy information [1575]. The critical knot revealed by these analyses is the enormous asymmetry in the perception of value: platforms derive colossal profits from combined profiling, but offer users marginal rewards (discounts, tokens, or fractions of a cent per view).

Here lies the ethical danger. If, to receive a few cents, a person is pushed to accept continuous profiling, which allows AI to craft spots created to exploit their emotional vulnerabilities of the moment, payment transforms from a "reward" into a coercive incentive to accept a totally unbalanced power relationship.

4. The European Shield: The EDPB and the "Consent or Pay" Model

The European legal system has already intercepted this distortion. When platforms like Meta introduced the "Pay or Consent" model (pay a subscription or accept advertising profiling), the European Data Protection Board (EDPB) intervened with clear opinions, which would also apply to a hypothetical model in which it is the user who is paid.

In its Opinion 08/2024, the EDPB establishes that "Consent or Pay" models must offer a real choice [1588, 1591]. For large online platforms (the so-called gatekeepers), offering the crossroads between total consent to behavioral advertising and a fee is almost never sufficient. Consent, to be valid under the GDPR, must remain free, specific, informed, and unambiguous [1593].

The EDPB suggests that, to remedy the power imbalance between user and tech giant, platforms must offer a free alternative that does not involve invasive behavioral profiling. The fundamental right to data protection cannot be transformed into a luxury good for those who can pay the subscription, nor can it be "bought" by platforms by taking advantage of users' economic need.

The Three Paradigms of the Attention Economy

To clarify, we can divide the evolution of digital advertising into three distinct models, each with its own exchanges and its own risks:

  1. Traditional Advertising (Broadcast):
    • What the user receives: Free access to the service.
    • What they give up: Generic attention, not individually tracked.
    • Key risk: Advertising saturation (being flooded with irrelevant spots).
  2. Behavioral Targeting (The current model):
    • What the user receives: Free access and more relevant ads.
    • What they give up: A continuous stream of personal data and the creation of a shadow profile.
    • Key risk: Hidden profiling, exploitation of cognitive biases.
  3. Remunerated Attention (The hyper-targeted horizon):
    • What the user receives: Free access and explicit compensation (money, discounts, tokens).
    • What they give up: Time, contractualized consent, absolute biometric/behavioral tracking.
    • Key risk: Economically conditioned consent, in which one accepts generative manipulation just to monetize one's time.

Key Operational Takeaways

  • Reject the Equation Money=Freedom: Consumers and legislators must reject the narrative according to which paying the user for their data "solves" the privacy problem. The surrender of fundamental civil rights (such as confidentiality) in exchange for micro-compensation is a form of coercion, not a free market.
  • Demand the "Non-Personalized" Alternative: As indicated by EU guidelines, users must systematically demand (and companies must provide) an option to enjoy services with purely contextual advertising (e.g., I see a shoe ad because I am reading an article about running, not because the AI has read my chats about the marathon).
  • Transparency on the "Generative Recipe": If a video is generated by AI in real time, the user must have the right to know which personal data points were used to craft it (e.g., "This video has anxious tones because we detected your recent searches about financial problems").

Conclusions: The Value of Our Free Will

We have gone from being spectators of a global shop window to being the product itself, and now we risk becoming pieceworkers of our own profiling. Generative Artificial Intelligence is the perfect engine for this transition: it does not limit itself to guessing what we like, but instantly creates the perfect stimulus to bypass our rational defenses.

The prospect of receiving micro-transactions to watch advertising modeled on our traumas, desires, and habits may seem the last frontier of market efficiency. But it hides an ethical question that our democracies will have to answer quickly: if an algorithm produces the emotionally most effective spot for me and a platform pays me to watch it, am I finally being fairly remunerated for my time, or am I selling – at a price decided by others – my very ability to choose freely?

Bibliographic References and Sources

Article by the Editorial Team of La Bussola dell’IA