EU Mission "A Soil Deal for Europe" — 100 Living Labs and Lighthouses by 2030; the 2026-2027 call structure and the €9m AI-ready soil decision-support topic

Europe (EU member states + associated countries; Living Labs organised by biogeographical region)

Content

The July 2026 funder scan named Mission Soil as a substantive unit candidate with a gap for its AI content (G-202). This unit closes that gap and corrects a figure: the AI content is one named topic of roughly forty million euros of 2026 mission money — €9m for AI-ready soil decision support — and the 2025 two-stage call was €48m, not the €36m the July scan recorded.

Headline. Mission Soil is a 100 Living Labs and Lighthouses by 2030 deployment mission, funded from the Horizon Europe Missions work programme. In 2026 it spends €81m across two call packages, of which one topic — HORIZON-MISS-2026-05-SOIL-04, “Leveraging long-term field experiments and other datasets to develop AI-ready decision support systems for sustainable soil management” (€9.00m) — is the AI-labelled instrument. No EU source publishes how many of the 100 Living Labs exist (G-405): only a funded-project roster of roughly twenty multi-site projects is verifiable.

The mission, verified 2026-09

“A Soil Deal for Europe: 100 living labs and lighthouses to lead the transition towards healthy soils by 2030.” One of five EU Missions under Horizon Europe, launched 2021. The Mission Soil platform states the same target: “the Mission Soil will put in place a network of 100 Living Labs and Lighthouses in rural and urban areas by 2030.” Living Labs are the mission’s deployment model — real-world co-design sites organised by biogeographical region, each required to run a multi-actor approach (farmers, researchers, businesses, civil society, public authorities).

Support structure: SOILL-Stepup (Grant Agreement 101262100, HORIZON-CSA, topic HORIZON-MISS-2025-05-SOIL-SGA-02), described in CORDIS as “the continuation of the SOILL Framework Partnership following the initial SOILL-Startup project”, coordinated by the European Network of Living Labs (ENoLL), last updated 27 March 2026. Its objective includes developing “the SOILL Hub … into a fully integrated collaborative platform offering information, training, monitoring and matchmaking” plus a multilingual helpdesk.

The 2026 call structure

Single-stage, HORIZON-MISS-2026-05 — opened 4 February 2026, deadline 23 September 2026, €33.00m total:

Two-stage, HORIZON-MISS-2026-05-two-stage — opened 12 February 2026, stage 1 closed 14 April 2026, stage 2 deadline 15 September 2026, €48.00m total:

• SOIL-01-two-stage: Living labs to enhance soil health in Alpine and Atlantic biogeographical regions — RIA, €24.00m (≈€12.00m × 2).
• SOIL-02-two-stage: Living labs to enhance soil health in managed forests and in natural/semi-natural lands — RIA, €24.00m (≈€12.00m × 2).

2027 is substantially larger: HORIZON-MISS-2027-05-two-stage (€94.80m, six topics including Continental/Black Sea/Pannonian/Steppic Living Labs at €24m and a eutrophication-reduction Living Lab cluster at €24m), plus the joint HORIZON-MISS-2027-06-SOIL-CANCER call on monitoring carcinogenic substances in soils (€24.00m, deadline 21 September 2027). Stakeholders meet at European Mission Soil Week, Dublin, 17-18 November 2026 (fourth edition, under the Irish Presidency).

SOIL-04 — what the AI money actually buys

The topic’s own text is the corpus’s clearest EU statement of what AI in soil management is for. It funds “AI-powered decision support systems”, requires the multi-actor approach, and — the operationally interesting part — permits financial support to third parties so that “beneficiaries may provide financial support to third parties to involve developers in creating or improving AI-powered applications that provide tailored advice to farmers and advisors… The maximum amount to be granted to each third party is EUR 60 000”. Expected outcomes include “enhanced adoption of impactful sustainable soil management solutions and strategies supported by AI-powered decision support systems by land managers” and a harmonised “logical architecture for long-term field experiments (LTEs) data collection and integration”.

The companion SOIL-01 (€5.00m CSA) funds the measurement layer underneath: “equipping stakeholders to sample, analyse, and interpret soil health indicators”. Read together, the 2026 package is soil measurement standards plus an AI decision-support layer over long-term field-experiment data — infrastructure for soil AI rather than soil AI products.

The project roster (not a count of Living Labs)

That is roughly twenty funded projects, each typically hosting multiple Living Lab sites — not twenty of one hundred Living Labs. Several are sensor- and remote-sensing-heavy (HydroDataCube4EU, LandPrint, LILAS4SOILS, CREDIBLE 2.0), and their AI content is documented through the CRCF and soil-MRV layer (units/eu-carbon-removal-certification-framework.md).

A caution on the July scan’s 2025 figures (subagent misattribution caught)

The July scan recorded the 2025 calls as €12m (single-stage, HORIZON-MISS-2025-05-SOIL-01, deadline 30 September 2025) plus €36m indicative for the 2025 two-stage call. A research pass this cycle read “48 million Overall indicative budget” and “33 million Overall indicative budget” off the REA Soil Mission page and treated them as corrections to those 2025 figures. That reading is wrong. €48.00m is the 2026 two-stage package (€24m × 2) and €33.00m is the 2026 single-stage package (€5m + €14m + €5m + €9m) — both read directly off the 2026-2027 work programme and internally consistent with its per-topic budgets. The 2025 figures were not re-verified this cycle either way (G-409); do not cite the 2026 totals as 2025 budgets, and do not cite the 2025 pair without re-checking the 2025 work programme.

The claim that the mission “will fall short of the 100-Living-Lab goal” remains an unverifiable judgement call: no EU source publishes progress against the target (G-405). What is verifiable is the funded-project roster and the deployment pattern.

What this unit is doing in the taxonomy

Anchors the EU mission-as-deployment-model layer: real-world co-design sites organised by biogeographical region, with a multi-actor requirement, distinct from both research consortia and vendor deployments. Distinguishes from:

• Horizon Europe Cluster 6 (units/horizon-europe-cluster-6-food-bioeconomy-agriculture.md) — the parent programme that contributes to this work programme part; Cluster 6’s own AI topics are research-and-innovation actions, whereas Mission Soil’s AI topic is an innovation action with third-party support.
• CRCF (units/eu-carbon-removal-certification-framework.md) — the certification framework that turns some Living-Lab soil outcomes into tradable units; the mission is the supply side, the CRCF the market frame.
• EAFRD / CAP (units/eafrd-cap-strategic-plans-digital-agriculture.md) — farm-level funding through member states; Mission Soil is EU-level and project-based.

Why it matters for talks

• The deployment model is the point. The EU’s preferred architecture for soil AI is the Living Lab — a co-design site with a multi-actor mandate — not a vendor product or a national platform. That is structurally distinct from US vendor-led and Chinese state-led deployment, and it is the corpus’s best example of the EU’s non-vendor-led posture.
• €9m for AI-ready decision support is the honest number. Against a €81m 2026 mission package and a €94.8m 2027 package, AI is one instrument, not the mission’s purpose. Useful counterweight to “the EU is building agricultural AI” framings.
• The €60,000 third-party cap is a concrete mechanism for getting small developers and local advisors into EU-funded AI applications — worth naming against the corpus’s recurring finding that smallholders and farm advisors are absent from agritech AI design.
• The unverifiable count is itself a finding. A mission with a 2030 headline target of 100 Living Labs publishes no running count; talks should say “roughly twenty funded projects, no published progress against the target” rather than repeat the 100 figure as though it were current.
• Long-term field experiments are the data asset. SOIL-04’s focus on LTE data harmonisation is the corpus’s clearest case of public-interest data infrastructure being built for AI in agriculture rather than extracted from farmers for it.

Critical context

• No published count of established Living Labs or Lighthouses (G-405); no per-project award amounts for the 2023-2025 Living-Lab projects verified from primary sources (G-406).
• Per-project AI content of the 2023-2025 Living Labs is documented only patchily; the 2026 AI topic is the first explicit AI-labelled mission soil topic.
• The mission’s timeline claim (100 by 2030) should be treated as a target, not a projection with evidence behind it either way.
• Support-structure details (SOILL Hub helpdesk launch, World Environment Day 2025) rest on secondary sources in this pass — flagged as unconfirmed-primary.
• 2026 deadlines are close: 15 September 2026 (two-stage stage 2) and 23 September 2026 (single-stage, including SOIL-04).