The printing press made knowledge accessible and triggered a century of religious wars. Radio brought continents closer together and served as a megaphone for totalitarianisms. The internet connected billions of minds while entrusting a significant share of that connection to a few large companies. Alicia Wanless, in The Information Animal, traces this long trajectory to pose a disturbing thesis: information technologies can broaden access to knowledge while opening new possibilities for capture, and the architecture of systems weighs heavily on the distribution of power.

The Essential Points

  • Major information technologies have often expanded access to knowledge while creating new control structures.
  • Wanless invites us to examine platform architecture as much as their contents to understand the organization of attention and opinion flows.
  • The European Digital Services Act, adopted in 2022, provides researchers access to certain data from very large platforms and search engines to study systemic risks; it does not, however, resolve the question of interoperability.
  • The central tension in the book opposes regulation of information environments to the preservation of innovations that make them useful.
  • Wanless underscores the possible interest of open standards for a more pluralistic information ecosystem.

A Researcher Between Two Worlds

Alicia Wanless directs the Information Environment program at the Carnegie Endowment for International Peace. Neither a media sociologist nor a designer, she occupies a rare space: that of a strategist who thinks of information ecosystems as terrains of influence, with the rigor of an analyst and the distance of an observer of information conflicts. Her previous work focused on disinformation and influence operations; The Information Animal broadens the framework to a long history of the relationship between humans and their communication technologies.

The Thesis: We Are Information Animals, and Always Have Been

The title is not ornamental metaphor. Wanless starts from an anthropological observation: Homo sapiens is a species defined by its capacity to process, store, and transmit information at a scale without equivalent in the animal kingdom. Our technologies are not external appendages to this nature: they are its extension, its amplifier, sometimes its prosthesis.

From this follows the central argument. When an information technology emerges, it can expand the possibilities for producing, circulating, and consuming information. At the same time, control of its infrastructure can weigh on the conditions under which this information circulates. The broadening of access and the concentration of control thus do not necessarily follow separate sequences: they can proceed from linked dynamics.

Wanless proposes reading this tension as a “dual trajectory” of information revolutions. Gutenberg’s printing press is the textbook case. Within a century, it multiplied the number of available books in Europe by a considerable factor, removed the clergy’s monopoly on the sacred text, and made the Protestant Reformation possible. Simultaneously, it created economies of scale that favored the concentration of the typographic industry, then enabled absolutist states to make it an instrument of censorship and propaganda. Access and capture were born from the same movement.

From Gutenberg to Algorithms: The Acceleration of the Cycle

What changes with digital platforms, according to Wanless, is less the nature of the phenomenon than its speed and scale. Dynamics formerly associated with long time frames can now unfold much more rapidly. Large platforms today occupy an important place in attention flows, particularly in many high-income countries.

But the most original point of the book is not there. Wanless refuses the easy explanation that reduces this phenomenon to platform greed or cynical exploitation of cognitive biases. She points to architecture. Recommendation algorithms are not technically neutral: they rest on objectives and ranking signals. Depending on the platforms, these can include engagement, viewing time, predicted interest, satisfaction, safety, or quality.

Engagement metrics can favor certain emotionally salient content, but they do not necessarily reduce to emotion and often coexist with metrics of satisfaction, relevance, or safety. This design choice, often presented as purely technical, is in reality a political choice about what behaviors the information environment produces collectively.

This reading aligns with a concern that work on AI and access inequalities is beginning to document: tools that promise to broaden access embed architectural choices that distribute gains very unequally. For Wanless, the mechanism is even more fundamental: inequality concerns not only who benefits from the tool, but who controls the rules of the environment in which the tool operates.

The Blind Spots the Book Leaves Open

The Information Animal is a book of diagnosis more than prescription. Wanless shows the possible limits of approaches exclusively focused on content, moderation, fact-checking, combating disinformation, when the structural conditions for information circulation remain unchanged. She is less convincing when it comes to saying how to touch this architecture without producing undesirable effects.

Economist Jean Tirole, whose work on market design and platform regulation has informed part of the European framework, points to exactly this tension: transparency can illuminate how platforms function without, by itself, changing competitive or dependency conditions. A regulator can force Meta to publish the criteria of its recommendation algorithms; the effect of such publication will also depend on the capacity of other services to offer viable alternatives. Transparency is a useful condition, but it does not automatically constitute structural reform.

Wanless mentions the interest of open standards, but the book does not seriously explore the obstacles to such openness. The definition of standards and their legitimacy remain central questions. Transitions to open standards in other technology industries—internet protocols, file formats, and payment interoperability—show that such changes require time and international coordination difficult to obtain in the platform field.

A competing, more skeptical reading regarding states’ capacity to regulate without capture deserves to be confronted with the book. Daron Acemoglu, whose work on institutions and technological power documents how technology winners can influence regulatory processes, would suggest that any platform regulation risks being influenced by the platforms themselves. The European DSA remains exposed to this risk, and Wanless does not directly confront it.

The DSA: Achievements and Limitations

The European example merits attention, as it is both the most advanced case of putting Wanless’s insights into practice and the revealer of their limits. The Digital Services Act, adopted in 2022, became generally applicable on February 17, 2024. For the first designated very large platforms and very large search engines in April 2023, specific obligations already applied by the end of August 2023.

Its Article 40 provides researchers with access to certain data from very large platforms and very large search engines in order to study systemic risks and measures taken to reduce them. Access to non-public data is subject to an approval procedure and regulated conditions, whose harmonized procedures were specified by delegated regulation (EU) 2025/2050, which entered into force in 2025. Researchers meeting the DSA’s conditions can thus request access to certain data from VLOPs and VLOSEs, without the text creating a general right for any academic team to freely audit the recommendation systems of Meta, TikTok, or YouTube. This is a real advance: for two decades, these systems were complete black boxes.

But interoperability remains absent from the main text of the DSA. The possibilities of transferring a social network, its connections or interactions to a competing service remain limited and variable depending on platforms. A user leaving Facebook can export some of their data, but does not necessarily have a way to recreate their network of relationships identically on a competitor. This architecture can contribute to high exit costs and strengthen the positions of established platforms, independent of any information quality question. Wanless is right to point to architecture as a knot in the problem; the DSA addresses only part of it.

Attempts by public administrations to use their own AI systems moreover illustrate the same dilemma at another scale: building alternative tools is not enough if uses and data remain captured by existing platforms.

Governing Information Ecosystems Without Stifling Them: The Question of the Decade

The Information Animal raises, without formulating it explicitly, a central question for the coming decade: regulating information environments in a way that preserves their capacity for innovation while neutralizing their tendency toward capture remains a challenge without an established answer.

Two trajectories are emerging, and current signals do not yet allow us to determine which will prevail. In the first, regulation remains centered on content and abusive behavior: moderation, transparency obligations, occasional sanctions against the most flagrant abuses. This path can leave platform architecture largely intact and preserve some of its structural advantages. Concentration can then persist, while economies of scale linked to generative AI develop, and states regulate behavior without necessarily touching the rules of the game. The DSA in its current state fits more into this trajectory than the other.

In the second, regulation would seek to act more on certain structural features: data portability, interoperability requirements in specific domains, or resort to open standards. This path exists as a proposal in several academic reports and in the work of a few pioneering regulators. It is more difficult to implement, because it requires international coordination that neither Europe alone nor the United States under any recent administration has been able to produce. It is also riskier: forcing interoperability between systems designed independently creates real security vulnerabilities, and premature standardization can freeze architectures that would have evolved.

Between these two trajectories, a third space emerges, more probable in the short term: geopolitical fragmentation of information environments. Europe imposes its rules on its domestic platforms and foreign actors operating on its territory; the United States oscillates between sectoral regulation and laissez-faire according to political cycles; China has long built an information ecosystem closed to its own rules. In this fragmented scenario, standards are not open but multiple, each geopolitical zone imposing its own. For users in middle-income countries, who have no means to influence any of these blocs, it is the worst of scenarios: they suffer rules from architectures designed elsewhere, without a voice.

What Wanless brings to this forward-looking reflection is less a roadmap than an analytical framework. Reading together the broadening of access and the concentration of control makes it possible to escape the temptation to personalize the problem. Regulators who think they can settle the question by sanctioning the behavior of Mark Zuckerberg or Elon Musk risk missing part of the structure. Comparable dynamics can appear when an information technology becomes central to social organization. They emerged with mass print in the nineteenth century, with radio in the 1930s, with television in the 1960s.

They could also manifest themselves with generative AI, whose development is accompanied by new information economies of scale and debates about the concentration of computing capacity, data, and models.

The Book’s Value

The Information Animal addresses three types of readers. To regulators and policymakers, it offers a historical framework that puts urgent responses centered on content in perspective and replaces the architectural issue at the center. To researchers in information sciences and digital economics, it proposes a unifying hypothesis, the dual trajectory, that would merit being tested empirically on more numerous historical cases. To cultured readers concerned with the governance of attention without having time to follow the technical literature, it provides a rigorous and readable entry point.

The book does not claim to solve what it analyzes. On the concrete modalities of implementing open standards, on the mechanisms of international governance that would make this implementation possible, on the arbitrations between innovation and regulation it would impose, Wanless holds back. This is both intellectual honesty—it maps a problem, it does not claim to have its solution—and a real limitation for the reader seeking actionable insights.

But perhaps posing the right problem is already, in this domain, a rare contribution. Most public debates about platforms revolve around content: disinformation, online hate, filter bubbles. Wanless shifts the view toward architecture, and this shift changes what one can see. Moderation policies attack what platforms let circulate; the question of open standards interests itself in the rules that organize this circulation. These are two different levels of intervention, and the confusion between the two partly explains why naming structural issues remains a rare act in public debate, whatever the domain.

Bibliographic Information Title: The Information Animal Author: Alicia Wanless Publisher: Hurst Publishers Publication date: May 2025

Sources

  1. Alicia Wanless, The Information Animal, Hurst Publishers, May 2025, https://www.hurstpublishers.com/book/the-information-animal/
  2. European Commission, Digital Services Act, https://digital-strategy.ec.europa.eu/en/policies/digital-services-act-package
  3. European Union, consolidated text of the Digital Services Act, including Article 40 on data access, https://eur-lex.europa.eu/eli/reg/2022/2065/oj?eliuri=eli%3Areg%3A2022%3A2065%3Aoj&locale=en
  4. European Commission, application of the DSA to the first set of very large online platforms, https://digital-strategy.ec.europa.eu/en/library/designation-decisions-first-set-very-large-online-platforms-vlops-and-very-large-online-search
  5. European Union, operational rules for accredited researchers’ access to non-public data, delegated regulation (EU) 2025/2050, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=OJ%3AL_202502050
  6. European Union, Digital Markets Act: portability and interoperability of messaging services, https://eur-lex.europa.eu/eli/reg/2022/1925/oj/eng
  7. Alicia Wanless et al., operational transparency framework, Carnegie Endowment for International Peace, https://carnegieendowment.org/files/202305-Operational_Reporting-final.pdf
  8. YouTube, official documentation on recommendations, https://blog.youtube/inside-youtube/on-youtubes-recommendation-system/
  9. TikTok, official documentation on recommendations, https://support.tiktok.com/en/using-tiktok/exploring-videos/how-tiktok-recommends-content
  10. Meta, Facebook data export tool, https://www.facebook.com/help/212802592074644?locale=en_GB