Copa-Cogeca — the EU farmer and agri-cooperative federation; 22 million farmers claimed, 15 EU projects, and no AI position of its own

Europe (EU-27 member organisations plus partner organisations in Iceland, Norway, Switzerland and Turkey)

Content

The July 2026 funder scan left Copa-Cogeca as a lead with a gap for its AI positions (G-203). This unit closes that gap as a negative finding: the EU’s primary farmer and cooperative federation has no stand-alone AI position paper, no AI policy area, and no locatable submission to the AI Office, the Code of Practice process or the Apply AI Strategy consultation. Its documented AI-adjacent position is a 2018 voluntary data-sharing code, and its operational AI engagement is one speaker slot on a Commission workshop agenda. The federation is a real institutional actor with a generational-legacy durability the corpus otherwise lacks — and on AI specifically it is largely absent.

Headline. Copa-Cogeca represents the farm-level constituency that all EU agrifood AI policy is meant to reach, holds a 2018 farm-data-sharing code that the Commission cites as the sectoral exemplar, runs 15 EU-funded projects of which none is AI-specific, and has published nothing on AI as such since 2018.

What it is, verified 2026-09

Copa-Cogeca is the united voice of EU farmers and agricultural cooperatives, headquartered in Brussels. Current site language: Copa “represents over 22 million European farmers and their family members”; Cogeca “represents the interests of 22,000 agri-cooperatives”. Copa has “60 organisations” from EU countries and “36 partner organisations” from other European countries including Iceland, Norway, Switzerland and Turkey, organised through 52 working parties. The secretariat is roughly 50 staff running 300+ meetings a year.

Founding dates — the July scan was wrong. Copa-Cogeca’s own timeline: Copa was created on 6 September 1958; Cogeca on 24 September 1959; Copa’s Brussels secretariat was established 1 April 1959 and merged with Cogeca’s on 1 December 1962. The corpus’s “Copa founded 1959” is off by a year.

The 22 million figure is Copa’s own claim and internally inconsistent. Copa’s 2018 joint publication states “Copa represents over 13 million farmers and their families whilst Cogeca represents the interests of 22,000 agricultural cooperatives. They have 66 member organisations.” The 22m number is also contested externally as a workforce statistic rather than a membership count. Use it as a claimed figure with the 13m/22m spread visible.

Leadership (September 2026): Copa President Massimiliano Giansanti (Italy); Cogeca President Lennart Nilsson (Sweden); Secretary General Elli Tsiforou (since September 2024); Deputy Secretary General Patrick Pagani; Chief Policy Advisor Paulo Gouveia.

The AI question — answered negatively

Copa-Cogeca’s policy areas do not include AI; AI sits inside digitalisation, on a “smart / precision farming” page. That page is the federation’s only standing AI-adjacent text, and it is framed as a farmer-interest position on data and transition, not a position on AI:

“Farming processes have become increasingly automated, connected and integrated. Data is now shared at an unprecedented level, which raises questions of ownership, access, control and security.”

“An ‘agri-tech revolution’ is impossible without farmers and agri-cooperatives.”

“We need a coherent strategy to promote the digital and technological transformation of European agriculture… All European policies, be they on the digital market, skills, infrastructure, investment, research and agriculture, must be aligned…”

The federation’s flagship data instrument is the EU Code of Conduct on Agricultural Data Sharing by Contractual Agreement (2018), published jointly by Copa-Cogeca and CEMA with Fertilizers Europe, CEETTAR, CEJA, ECPA, EFFAB, FEFAC and ESA (the European Seed Association). It is explicitly “non-binding guidelines”. Its core construction: the operator producing data is “the data originator”; “the data originator must give permission for their data to be used and shared with third parties, including circumstances in which decisions are made based on the data”; and “data must not be used to assess the originator’s ability to pay for a service or machine”. The Commission cites it as the sectoral exemplar in its 2021 AI Communication.

The negative findings are the finding. No stand-alone AI position paper. No AI policy area. No verified Copa-Cogeca or CEMA entry among the 287 feedback entries on the Apply AI Strategy call for evidence (recorded as not found, not as non-submission — the register’s keyword filter did not work and only a sample was inspected). No participation in the AI Act Code of Practice process found. And, per this cycle’s AI Office verification, no agrifood or farmer organisation at all sits on the AI Act Advisory Forum (units/european-ai-office-governance-architecture.md). The one verifiable AI-adjacent engagement is operational: the DG AGRI + DG CNECT expert workshop “Fostering AI uptake and scaling trusted AI solutions in agriculture” (24 June 2026), whose agenda lists Francesca Hennig-Possenti as Chair of CEMA’s AI Group — and names no Copa-Cogeca speaker.

The project portfolio — 15 projects, none AI-specific

The portfolio (per copa-cogeca.eu/projects) is knowledge, skills, safety and advisory work. The AI-adjacent entries are QuantiFarm (July 2022 – December 2026; “support the increased deployment of digital technologies in agriculture… data-driven technology innovations”) and the completed SmartAgriHubs (September 2018 – October 2022, which claimed “80 new digital solutions” and helping “digitise over 2 million farms”), plus EU-FARMBOOK (August 2022 – July 2029, an EU-wide digital platform for AKIS knowledge exchange) and EU4Advice (October 2022 – September 2027).

Corrections to the July scan’s list: Smartchain (2019-2021), Innoseta (2018-2021) and GAEA (2022-2024) are completed, not current; Social Dialogue for Skills (SD4S, September 2023 – September 2025) and Growing the Future (January 2020 – June 2021) were missing from the corpus list. There are exactly 15 projects, not “15+”. Per-project AI deployment outcomes are not published for any of them (G-411).

The European Space Agency partnership

Copa-Cogeca and ESA signed a first joint Memorandum of Intent, “Supporting the Sustainable Transformation of Agriculture and the Agri-food Sectors through Digitalisation and Space”, in Paris on 28 September 2023. Christiane Lambert (then Copa President): “The use of digital and space technologies is already key for many of us, but it will become increasingly important for the strategic autonomy required by the European Union.” ESA’s Nick Appleyard: “Copa-Cogeca are the perfect partner to accelerate the green transition of the agri-food sector.”

Named activity: ESA opened the ARTES BASS call “The Sustainable Transformation of Agriculture Through Digitalisation and Space” on 13 October 2023, closing 15 January 2024, covering topics including “Agriculture of Data” and “edge-computing and AI capabilities” in Farm Management Information Systems; Patrick Pagani of Copa-Cogeca spoke at the 25 October 2023 webinar. No funded demonstrations or deployed services under this partnership are published (G-410) — the partnership is a memorandum and a call, not yet a deployment record.

What changed since July 2026

Nothing on AI. Copa-Cogeca’s post-July output is entirely conventional agricultural policy: CBAM fertilisers (7 July 2026), fertiliser-crisis CAP adjustments (12 June 2026), NGT final EP approval (17 June 2026), Nitrates Directive evaluation (15 July 2026), public procurement (9 September 2026), IED livestock opinion (2 September 2026), a heatwave assessment to Commissioner Hansen (1 September 2026), cereal forecasts (23 July 2026) and milk-and-dairy working-party elections (8 September 2026). As of 14 September 2026, the 2018 code remains the federation’s most substantive AI-adjacent instrument.

What this unit is doing in the taxonomy

Anchors the EU farm-voice institutional layer — the constituency-side counterpart to the funder and regulator layers. Distinguishes from:

• CEMA (units/cema-eu-agricultural-machinery-association.md) — the equipment-maker federation, which does take documented AI positions and co-published the 2018 code.
• Mondragón (units/mondragon-corporation-cooperative-federation.md) and COVAP (units/spain-cooperative-covap-ai-deployment.md) — cooperative operators deploying AI, as against a representative federation.
• EIT Food (units/eit-food-eu-knowledge-innovation-community.md) — institutional funding of startups; Copa-Cogeca sits on the farmer side of the same EU architecture.
• European AI Office architecture (units/european-ai-office-governance-architecture.md) — where the absence of a farmer seat is documented.

Why it matters for talks

• The absence is the talk line. The EU’s largest farm federation has no AI position and no AI governance seat, while the EU regulates AI and funds agrifood AI. “Farmers are at the centre of EU agrifood AI policy” is not supportable from the institutional record — the farmer voice in EU AI governance is currently zero seats.
• The 2018 code is the corpus’s clearest farmer-data-rights instrument in Europe — and it is non-binding, voluntary, and eight years old. Pair it with JoinData (units/joindata-netherlands.md) as the contrast between a voluntary code and an actual data cooperative.
• “An ‘agri-tech revolution’ is impossible without farmers and agri-cooperatives” is the federation’s own best line and the corpus’s cleanest expression of the constituency claim — quotable against vendor-led deployment framings.
• The 13m-versus-22m membership spread is a small, concrete demonstration of the corpus’s verification discipline: an institutional claim that does not survive its own archive.
• QuantiFarm and SmartAgriHubs are the two EU digital-agriculture programmes where the federation has real standing — useful when a talk needs EU digital-adoption projects with an institutional partner rather than a vendor.

Critical context

• 22 million is Copa’s own current claim; its 2018 publication said over 13 million, and external investigations dispute the figure as a workforce rather than membership statistic.
• “15+” projects is exactly 15; three the July scan implied were current are completed.
• No AI position paper, no AI policy area, no verified AI consultation submission, no Code of Practice participation (all recorded as not-found, not as proven non-participation).
• ESA partnership: memorandum and call only; no funded demonstrations published.
• Per-project AI deployment outcomes are unpublished across the whole portfolio, so “Copa-Cogeca projects deploy AI” should not be asserted from QuantiFarm’s digital-technology framing alone.