An artificial intelligence that does not merely compete, but serves. That, in the shortest possible terms, is the proposition behind AI AnthropoCosmic, the international project promoted by the Faculty of Social Communication Sciences of the Pontifical Salesian University in Rome, led by its Dean, Prof. Fabio Pasqualetti, and presented in the press release of 10 July 2026. The official payoff is unambiguous: “For an AI at the service of Humanity, the Environment and the Cosmos. Without knowledge, the distance between technology and life is an abyss.” This is not, however, a rhetorical exercise: the project sets itself an avowedly operational objective, namely to “contribute to the definition of shared guidelines capable of orienting public policies, technical standards and industrial models”. And anyone who works with technical standards and compliance knows how much weight that one word, standards, carries in such a list.
I am honoured to be part, together with AI Open Mind, of the team supporting the development and the practical implementation of this initiative.
What AI AnthropoCosmic is
The project starts from a diagnosis: the currently prevailing model of AI development follows “a predatory and competitive logic”. The alternative it proposes is a “collaborative model founded on knowledge, responsibility and sustainability”, one that treats AI as a “global common good”. The clarification, a welcome one, is that “the objective is not to slow down technological innovation, but to orient it towards the common good”: not a brake, then, but a direction.
Around the Pontifical Salesian University a group of actors has taken shape which deserves careful reading, because it says a great deal about the character of the initiative. The operational partner is AIopenmind, a non-profit association that has built the project website and hosts the participation forms. Institutional patronage comes from UNINFO, the UNI federated body responsible for information technologies. The media partnership is with Media Duemila. This is no small detail: the project expressly declares its continuity with the work begun on 3 May 2024 within Technical Committee UNI/CT 533 at UNINFO on “AI Act, Data Act, Data Governance and Standards”. The thread connecting humanistic inspiration to technical standardisation is, in other words, explicit.
The three dimensions: Humanity, Environment, Cosmos
The conceptual architecture unfolds across three concentric levels. The first, Humanity and AI, concerns the centrality of the person, human rights, critical education, work and democracy. The second, Humanity, AI and the Environment, addresses sustainability, biodiversity and the energy impact of models, a question the legal debate has so far approached with a certain reticence. The third, Humanity, AI and the Cosmos, looks to cooperation between States, global governance, AI deployed in space and quantum computing.
The eighteen “AI Laboratories of Tomorrow”
The most interesting element, for those who work with standards, is the structure of the “AI Laboratories of Tomorrow”: eighteen sectoral ecosystems drawn from Technical Report ISO/IEC TR 24030:2024 “Information technology – Artificial intelligence (AI) – Use cases”, mapped by Domenico Natale, honorary member of UNINFO. Anchoring the working groups to an ISO/IEC document, rather than to a taxonomy invented for the occasion, is a methodologically sober choice and makes contributions comparable. Among the eighteen ecosystems is the legal sector: for once, the legal profession is not treated as a bystander. Participation in the Laboratories is arranged through the dedicated form.
The path: three seminars, a conference, a White Paper
The path unfolds along four routes of participation, with firm deadlines. Three free online seminars mirror the three dimensions: “Humanity and AI” on Saturday 7 November 2026, “Humanity, AI and the Environment” on Saturday 14 November 2026, and “Humanity, AI and the Cosmos” on Saturday 21 November 2026. The international conference “AI Anthropocosmic: per un’IA a servizio dell’uomo, dell’ambiente e del cosmo” will be held on Saturday 28 November 2026, in person and online, in the Aula Paolo VI, Piazza dell’Ateneo Salesiano 1, Rome, and will produce a final document. Registration for the seminars and the conference is open.
The fourth route is the Call for Paper for the White Paper on Artificial Intelligence, and it carries the nearest deadline: 15 September 2026. Contributions may run to a maximum of 3,500 characters, accompanied by an abstract of no more than 500 words, and must be consistent with one or more of the eighteen Laboratories. Papers are assessed by a Scientific Committee; those approved will be presented at the seminars and the conference and published in the collection “AI Anthropocosmic – AI Laboratori del Domani”. Submissions are made through the dedicated form, with the Word file sent to fsc.aiopenmind@gmail.com. The initiative’s first echoes are collected in the press review, which includes articles by ADVexpress (10 July 2026), Media 2000 (12 July) and Cavaliere News, signed by Diana Daneluz. On the origins and the cultural framework of the project, see also the presentation published by the Faculty of Social Communication Sciences.
What a White Paper can do, and what it cannot
It is worth being precise about effects. A White Paper is not a technical standard and produces no presumption of conformity: it gathers positions, submits them to the assessment of a scientific committee and organises them. Its value lies upstream, at the stage where the debate is still open and the content is still malleable. The channel through which those positions may eventually enter the standardisation circuit is declared by the project itself: continuity with the work started in the Technical Committee UNI/CT 533 at UNINFO. It is there, and in the European and international bodies, that ideas become requirements.
For those who work on law and technical standardisation, therefore, the interest of the initiative does not lie in the statement of principle, but in the method: anchoring the working groups to an ISO/IEC document rather than to a taxonomy invented for the occasion, submitting contributions to assessment, setting deadlines. One wonders whether this will be enough to genuinely orient public policy, technical standards and industrial models: the answer will depend, as always, on the quality of the contributions that arrive by 15 September.
