Magnifica Humanitas: What Pope Leo XIV’s Encyclical Tells Us About Digital Service Taxes

Pope Leo XIV’s Magnifica Humanitas, published on the 135th anniversary of Rerum Novarum, calls for data to be treated as a common good. It denounces a new form of digital colonialism, where user data is harvested without compensation. This resonates very much with the policy rationale that has driven the introduction of Digital Service Taxes (DSTs) in a growing number of jurisdictions. While Amount A of the OECD’s Pillar One was designed to reallocate taxing rights over the excess profits of large digital companies to market jurisdictions, it remained within the boundaries of existing concepts and tried to update them rather than exploring solutions that would focus on the role of users’ data in the value chain of certain highly digitalized business models.

DSTs were conceived as a response to the stall at multilateral level and as an alternative approach to capture the value created by users in what was then one of the core features of the Web 2.0, namely, the platform economy. Today we are in the middle of another and more powerful transition, the one of the AI economy.

This passage of the Encyclic attracted my attention:

Just laws and methods of redistribution are certainly necessary for correcting imbalances, including tax systems that lighten the burden on the weakest and ask for more from those with greater resources. However, the pursuit of social justice should not be considered a separate issue that follows only after the production of wealth, as if the economy existed solely to create wealth, with politicians only intervening afterwards in order to distribute it. Indeed, justice concerns every phase of economic activity, from resource acquisition to financing, and from production to consumption; every choice has moral consequences (162).

It made me think about the potential evolution of DSTs in the AI economy, their rationale (tax the value generated by users’ contribution), proxies (revenue generated from key data monetization activities), and rates (flat at the moment).

There may in fact be a case for differentiated rates calibrated to the weight of the connection between the revenue generating activity and the degree of monetisation of users’ data. A rough hierarchy is actually rather identifiable under current DSTs. At the first level sits the direct sale of data, pretty much self-explanatory. At the second level sits targeted advertising, data are not sold but monetized via observation of the preferences of users and their profiling so that ads are more effective. At the third level currently sits digital intermediation, where data drive matching efficiency, indirect network effects, users’ retention and dynamic pricing.

Once revenues (the DST tax base) is considered to be a suitable proxy to identify data monetization, differential DST rates could be justified based on the above. This topic was discussed at 2025 University of Amsterdam Centre for Tax Law’s (ACTL) Conference Digital Service Taxes (DSTs): A fair and effective way to tax the data economy? Within the context of the CPT Project, the conference, gathering leading scholars, policymakers, and practitioners examined the legal, economic, and policy dimensions of these levies.

The chief question remains however which revenues should be in scope of a DST. The focus has so far been on the platform economy and typical monetization strategies. Today we are already squarely into the transition to the AI economy. While some of the key players are the same, there are also fundamental differences in the underlying business models.

At the infrastructure layer are the hyperscalers who provide the compute, storage and cloud capacity on which AI can function. The question here would be to what extent data collected across the ecosystem are a key element of the monetization strategy. At the model layer are the large language model developers, including  the AI divisions of the hyperscalers themselves, who train foundation models and distribute them via API. The question here would be whether, and to what extent, user-generated data and interaction feedback loops are central to the value of their models and their commercial proposition. At the application layer are the “wrappers”, the vertical applications built on top of foundation models, from coding copilots to legal assistants and customer service tools. The question here would be whether the collection of conversations, preferences and behavioral signals is incidental to the service or structural to the business model.

Interestingly, and differently from the posture adopted in the wake of the discussions on the taxation of the digital economy, key players in the AI space have been vocal on the need to regulate the new wave of change. The convergence of industry voices on this point is telling.

Chris Olah, co-founder of Anthropic, spoke at the Vatican presentation of Magnifica Humanitas and warned that AI development cannot be left solely to technology companies and urged greater oversight from governments, religious institutions, and civil society. Olah acknowledged that AI could displace human labor at very large scale, adding that supporting those displaced would become a moral imperative of historic proportions. He also noted that companies like his operate under intense commercial, geopolitical and personal pressure, a candid admission that the self-regulatory capacity of the industry has structural limits.

This resonates with Pope Leo XIV statement that:

More than ever, in the age of AI and robotics, it is no longer possible to rely solely on the “invisible hand” of the market. Politics has the task of orientating economies and technologies to the common good, promoting dignified work, social inclusion and an equitable distribution of the benefits of innovation. Since many economic decisions transcend national borders, there is also a need for international cooperation capable of defining common strategies, especially in favor of the most vulnerable countries and people, in order to promote development and overcome welfare dependency. The thinking behind these choices is the immeasurable dignity of every person, the common good and a world truly governed for everyone. The interdependence between peace and development, as Saint Paul VI prophetically wrote in 1967, remains applicable today, for prosperity contributes to building and reinforcing peace only if it is widespread, inclusive and sustainable (163).

Olah is not the only one who underlines the need for new approaches. Writing in the Financial Times, last month Arthur Mensch, co-founder and CEO of Mistral AI, made a proposal for an AI content levy in Europe. He argued that AI companies should give back a portion of their revenue to compensate authors and creators for the use of their work in model training, proposing a revenue-based levy applicable to all AI companies wishing to offer services in Europe, whether European or not. His stated motivation was partly competitive: Chinese firms, he argued, are training their models on vast amounts of European content with little regard for copyright rules. But the underlying logic is fiscally familiar — a levy calibrated to revenue, applied at the point of market access, designed to capture value that would otherwise be extracted without contribution to the jurisdiction in which it originates.

The parallel with the DST rationale is evident, even if the revenue destination differs: where DSTs route revenue to the general budget, Mensch’s levy would channel it to the cultural sector. In a way, the above shows that the conversation is already moving beyond the usual interlocutors, corporate tax experts who see through their highly technical glasses, NGOs who accuse large tech companies of dodging corporate taxes and politicians who are stuck in the middle.

DSTs are an imperfect instrument, as most taxes are. DSTs have been controversial since their inception and still are today. The February 2025 US Memorandum targeting them as “overseas extortion” added further geopolitical turbulence.  As discussed in a forthcoming book to be published by the CPT Project and Wolters Kluwer titled "Digital Services Taxes: National Experiences, Policy and Legal Aspects", the experience of countries that have implemented DSTs is on balance a positive one. Revenues, modest but not negligible, increase year by year, as the number of taxpayers that are in scope. No major operational issues have arisen and compliance is relatively straightforward, something that would probably be difficult to say about Pillar One, and the recent experience with the global minimum tax under Pillar Two shows just that.

Magnifica Humanitas reminds us that treating data as a common good is not a pious aspiration. It is actually a policy mandate.

… For this reason, it is essential that the use of AI, especially when it touches on public goods and fundamental rights, be guided by clear criteria and effective oversight, grounded in participation and subsidiarity. Communities and intermediary organizations must not be reduced to passive recipients of decisions made elsewhere; they must be able to contribute to discernment and oversight. Moreover, ownership of data cannot be left solely in private hands but must be appropriately regulated. Data is the product of many contributors and should not be treated as something to be sold off or entrusted to a select few. It is necessary to think creatively in order to manage data as a common or shared good, in a spirit of participation, as Saint John Paul II already suggested regarding collective goods (108).

Policymakers and experts are called to contribute on how to design a tax system that is more precise, more coherent, and more fit for the AI age. This relates to DSTs but also to wider topics such as the adequate tax mix in the AI age, the role of income taxes and the need for wealth/solidarity taxes. A small but hopefully useful contribution to the debate will be given at the conference that the University of Amsterdam ACTL is organizing on 12 and 13 November 2026 titled International Taxation in the New Global Order.

The purpose that Magnifica Humanitas has articulated is very clear: ensuring that the digital transformation of our economies does not become an inheritance of injustice for the generations that will come after us.

*Raffaele Russo is Partner at Chiomenti and Senior Fellow at the University of Amsterdam: Views are his owns.