MF · CSR & Green IT Sovereignty · 2026
Module MF — Digital Sovereignty
Sovereign &
responsible
digital.
AI carbon footprint · Green IT · CSRD · Sustainable strategic autonomy
CSR & Sovereignty convergence — ORBii Framework

"A non-EU cloud consuming energy produced outside the EU, in a datacenter beyond your control, with an opaque carbon footprint, is not just a sovereignty risk — it is also a CSRD risk."

What this module covers

AI model carbon footprint, environmental impact of cloud, measurement methodology, CSRD reporting, sovereign Green IT strategy.

Target audience

CDO, CIO, CSR/CSRD officers, cloud architecture teams, senior management, sustainable procurement managers.

Core regulation

CSRD (2024-2026), EU Taxonomy, datacenter energy efficiency directive (EU EED), ISO 14001, GHG Protocol Scope 3.

Recommended duration

Half day (3h30) — 2 sessions + 1 digital footprint measurement workshop.

ORBii.Academy — Digital Sovereignty & AIMF · P.01
MF · CSR & Green IT Sovereignty
AI carbon footprint

What AI models truly cost the planet.

Generative AI is a high-energy-intensity technology. Its carbon footprint directly depends on infrastructure choices — and therefore on sovereignty choices.

10x
A generative AI query consumes on average 10 times more energy than a standard web search
2-3%
Share of global electricity consumption attributable to datacenters — growing rapidly with AI
Scope 3
LLM consumption via external API is Scope 3 — subject to CSRD reporting from 2026 for large enterprises

The 3 footprint phases of an AI model

Phase 1 — Training

The most energy-intensive phase. Training a large model can consume as much as hundreds of transatlantic flights. This phase is beyond the reach of user organizations — unless they train their own models.

Phase 2 — Fine-tuning

If performed on own infrastructure (on-premise or EU cloud), the footprint is measurable and traceable. If performed via external API, the footprint is opaque and cannot be accurately reported in the CSRD statement.

Phase 3 — Inference

Every query to an external LLM generates a Scope 3 footprint. At the scale of a banking organization with thousands of AI tool users, the volume is significant and must be measured.

CSRD — What is mandatory from 2025-2026

Large companies subject to CSRD must report their digital footprint including AI usage in non-financial reporting. Any consumption via a cloud provider whose energy data is opaque makes the reporting non-auditable — and exposes the organization to a non-compliant qualification by statutory auditors.

ORBii.Academy — Digital Sovereignty & AIMF · P.02
Protected content

You have viewed the preview of this module (first 2 pages).
To access the full content, enter your access code or request access.

6 remaining pages Personal link · Valid 24h