Skip to content
Henri de GrossouvreEssayist on European geopolitics
04

Expertise

Sovereignty technologies & foresight

In 2005 he wrote that a Europe of power would be decided by the mastery of sovereignty technologies. In 2025 he co-signed a Figaro op-ed saying that data are the key to artificial intelligence. It is the same subject, twenty years apart.

The argument

This is the field where it would be easiest to overclaim. It therefore deserves to be set out precisely.

What the 2005 text says

In the October 2005 issue of Revue Défense Nationale, Henri de Grossouvre places Europe before a choice: a Europe of power built on a Carolingian core, or a Europe of markets reduced to a free-trade zone under someone else’s strategic protection. The journal’s own presentation of the article states that European leadership requires continental strategic cooperation in order to master sovereignty technologies.

That formulation is not ours. It is the journal’s, in 2005.

What it became

In October 2025 he co-signed, in Le Figaro, with thirty-four other signatories — Luc Ferry, Arnaud Montebourg, two former defence ministers, four generals, two senators — an op-ed on digital sovereignty. The text deals with French and European cloud dependence on providers subject to the extraterritorial reach of US law, with the “cloud at the centre” doctrine, and with ANSSI’s SecNumCloud certification. It casts data as the key to artificial intelligence. And its central analogy — the same ambition as for nuclear power in the 1960s — restates in the present tense what was written twenty years earlier.

What lies between the two

The continuity is not a retrospective reconstruction: it is marked out.

In 2007, in the same journal, he reported on the creation of the ID Advance grouping and the aerospace and transport industrial campus near Strasbourg. A piece of industrial policy written from inside. In 2021 he co-signed an article on opportunities for EU-China technological cooperation around decarbonisation. In 2024 and 2025 he published three columns: on deep learning applied to cities, where he holds the benefit to be conditional — privacy, data quality, compute cost, bias; on the legal framing of AI voice synthesis, starting from the Tennessee statute, and on the case for European harmonisation; on satellite data, technologies of military origin turned into planning instruments, whose obstacles he names rather than passing over.

What this field is not

Henri de Grossouvre is not an artificial-intelligence researcher and this site will never present him as one. He has no research corpus on AI, and claims none.

What he has is an argument set out in 2005 on the technological condition of European autonomy, an industrial practice that went with it, and a position published in 2025 that restates it in the present on the ground of data and compute. That is a different thing, and a sturdier one, because every link is dated and linked.

The question that stays open is the one the file poses: can Europe master artificial intelligence without mastering its energy, its networks and its data?

Key facts
  • The phrase “sovereignty technologies” appears in the journal's own presentation of his 2005 article in Revue Défense Nationale.
  • In 2007, in the same journal, he reported on the creation of the ID Advance grouping and the aerospace industrial campus near Strasbourg: the argument came with a practice.
  • In October 2025 he co-signed, with thirty-four other public figures, a Figaro op-ed on European cloud dependence on US law.
7

Related publications

ArticleOctober 2005

“Fortuna and a Europe of Power” (opens in a new tab)

Revue Défense Nationale, n° 679

The pivotal text in the corpus. Europe is set before a choice: a Europe of power built on a Carolingian core, or a Europe of markets reduced to a free-trade zone under someone else's strategic protection. The journal's own summary states that European leadership requires continental strategic cooperation in order to master sovereignty technologies. Twenty years before compute, models and data became exactly that.

Collective op-ed29 October 2025

Digital Sovereignty: “France Must Show the Same Ambition It Showed for Nuclear Power in the 1960s” (opens in a new tab)

Le Figaro, FigaroVox — 35 signatories

A collective op-ed signed by thirty-five public figures, Henri de Grossouvre among them, under the description “essayist specialising in international relations and European geopolitics”. The piece warns of French and European cloud dependence on providers subject to US law, discusses the “cloud at the centre” doctrine and SecNumCloud certification, and casts data as the key to artificial intelligence. Its central analogy — the same ambition as for nuclear power in the 1960s — restates in the present tense the argument set out in 2005 in Revue Défense Nationale.

ArticleMarch 2021

The New International Landscape and Opportunities for EU-China Strategic Cooperation (opens in a new tab)

Revue Défense Nationale, n° 838

A collective article published in the file “Have we entered the post-Western era?”. The authors argue that decarbonisation demands greater international cooperation, and that this offers the European Union and China an opening to move past their antagonisms and build a shared project.

With François Loos, Joël Ruet, Fatima Hadj, Alex Wang, Jean-Claude Beaujour
Op-ed18 April 2024

AI and Synthetic Voice: Could France Draw on the ELVIS Act? (opens in a new tab)

Journal du Net

Starting from the Tennessee statute protecting performers' voices against unauthorised synthesis, the author asks whether France should transpose it and whether Europe should harmonise. He sets out the tensions rather than resolving them: protection against creative freedom, the legal definition of “voice”, and the law's capacity to keep pace with the technology.

Op-ed31 January 2025

How Satellite Data Can Support Smart-City Solutions (opens in a new tab)

Journal du Net

Technologies born of military need turned into planning instruments: infrastructure monitoring, urban planning, natural-hazard management. The author names the obstacles — the cost of acquiring and processing data, inadequate computing infrastructure — rather than passing over them.