Canada deepening cycle — talk-ready findings (July 2026)

Canada deepening cycle — talk-ready findings (July 2026)

Status. Cycle-level consolidation of substantive findings for talk-building. The 18 new units each have their own “Why this matters for talks” section (load-bearing); the 3 scans each have their own “What this scan asks of the field guide” section (cross-unit synthesis). This document collapses those into a single navigable talk-ready index. Use this document when building a talk and you need to know which of the 18 units to pull from, and which of the 10 substantive findings are the headline arguments.

What changed in this cycle. The Canadian corpus previously had three Canada-side scans (FCC cycle, Quebec cycle, regional scan) with strong federal Crown corporation + Quebec institutional substrate, but zero dedicated units for the federal funder/convenor substrate (CAAIN, Scale AI, Digital, PIC, RAII), zero substantive value-chain matrix cells (animal production, mushroom, orchard, grain grading), and zero regional substrate units (AIDA, Prairies, BC, North, DFO, aquaculture). This cycle populates all three.

Archetype integration. The 10 substantive findings below slot into the five existing archetypes that cover Canada. Each finding maps to one or more archetypes. The Canada-specific adoption-diagnosis (archetype 03) and data-sovereignty (archetype 02) archetypes are the most directly upgraded; archetype 04 (cooperative-alternative) gains new Canadian anchor units; archetype 05 (critical-lens Indigenous sovereignty) gains G-313 / PolArctic / Salmon Vision evidence; archetype 06 (regional-cluster-comparison) gains a substantively-evidenced Canadian cluster pattern.

1. The five-funder federal substrate (headline for adoption-diagnosis)

The Canadian federal agrifood AI substrate in mid-2026 comprises five distinct federal channels, layered on the FCC-led Crown corporation substrate:

• CAAIN (Canadian Agri-Food Automation and Intelligence Network) — ISED SIF + AAFC Clean Tech, $19.1M+ CAAIN contribution against $52M+ total project value across 35+ projects.
• Scale AI (Global Innovation Cluster) — up to $284M federal, 13 named Agriculture & Mining projects.
• Digital / DIGITAL Technology Supercluster (BC-anchored) — multi-hundred-million, 20+ Natural Resources & Agriculture projects.
• Protein Industries Canada AI Programme — $30M PCAIS-delivered, ended March 2026, 7 projects.
• Regional Artificial Intelligence Initiative (RAII) — $200M Budget 2024 / $500M AI for All scaling, delivered by 7 RDAs.

This is federally orchestrated + cluster-channeled + regionally delivered — a third model worth naming in its own right, distinct from the EU’s DG-distributed substrate and the US private-venture substrate.

Talk-ready claims.

• “Canada has five federal channels for agrifood AI funding, layered on FCC’s Crown corporation substrate.”
• “CAAIN is the largest single substantive content surface in the corpus — $19M+ CAAIN contribution against $52M+ total project value across 35+ projects.”
• “The PIC AI Programme ended March 2026 with no successor named in AI for All (June 2026).”
• “RAII scales from $200M to $500M in AI for All (June 2026) — federal AI SME funding is scaling, not contracting.”

Anchor units. units/caain-portfolio-canada.md, units/scale-ai-agriculture-canada.md, units/digital-supercluster-agriculture-canada.md, units/protein-industries-canada-ai-programme.md, units/raii-canada-ai-adoption-programme.md, units/fcc-canada-ai-adoption.md.

Archetypes. 03 (adoption-diagnosis) — primary; 06 (regional-cluster-comparison) — secondary.

2. The cross-cluster funding pattern (headline for vendor-sweep)

At least 5 named leads (Vivid Machines, Terramera, Agi3, Verge Ag, EarthDaily Analytics) appear in both Scale AI and Digital Technology Supercluster. Same companies, multiple federal funding streams. The cluster programme architecture allows stacking.

Talk-ready claims.

• “Canadian cluster programmes are not siloed — at least 5 named leads stack federal funding across Scale AI and Digital.”
• “The same company can be CAAIN + Scale AI + Digital-funded (Vivid Machines: CAAIN Open Competition + Scale AI Transforming Fruit Production + Digital Apples to All).”
• “Federal cluster stacking materially changes deployment economics — this is a structural Canadian feature.”

Anchor units. units/scale-ai-agriculture-canada.md, units/digital-supercluster-agriculture-canada.md, units/canadian-orchard-ai.md (Vivid Machines triple-funded).

Archetypes. 06 (regional-cluster-comparison) — primary; 01 (vendor-sweep) — secondary.

3. The funding-stack pattern in aquaculture and the data-sovereignty architecture (headline for data-sovereignty)

The same kind of AI capability (computer-vision fish monitoring) shows up under each Canadian funding regime, but with different data-governance postures:

• Venture-funded (ReelData AI Halifax) — proprietary but customer-licensed.
• PCAIS-funded (OnDeck Fisheries AI Vancouver, $1.5M Ocean Supercluster) — public infrastructure.
• Corporate-funded (Mowi Canada West Feed Centre) — proprietary stacks.
• Cluster-funded (Ocean Supercluster H2S-sensor + Grieg Seafood fish welfare).
• NGO + DFO-context-funded (Oceans North Atlantic EM pilot).
• Indigenous-led (PolArctic Sanikiluaq Nunavut; Salmon Vision BC) — community-stewarded.

Substantive observation: Same CV-on-fish capability looks structurally different depending on which Canadian funding regime produced it. The regulatory-funder mix shapes the data-sovereignty architecture, not just the deployment.

Talk-ready claims.

• “Canadian aquaculture AI has six distinct funding regimes producing structurally different data-governance postures.”
• “The same computer-vision-on-fish capability looks structurally different depending on the funding source.”
• “Indigenous-led AI deployment (PolArctic, Salmon Vision) is operationally distinct from vendor-funded commercial AI — community-stewarded vs. proprietary.”

Anchor units. units/canadian-aquaculture-ai.md, units/northern-canada-can-ai-2026.md, units/bc-horticulture-ai-cluster.md, units/dfo-pacific-salmon-ai.md.

Archetypes. 02 (data-sovereignty) — primary; 05 (critical-lens) — secondary.

4. Animal production is now substantively populated for the first time (headline for vendor-sweep)

Pre-cycle, animal production was the thinnest cell in the Canadian corpus (only SoraLINK × Saputo/Olymel/Agropur as a processing-side dairy anchor). Post-cycle:

• Dairy (4 CAAIN projects + SoraLINK): CATTLEytics, SomaDetect, Milk Moovement, Lakeland Precision Ranching + SoraLINK.
• Meat processing (3 CAAIN projects): P&P Optica, mode40, Circulus Agtech.
• Poultry (3 CAAIN projects + Targan WingScan): Chick Pick, Farm Health Guardian, MatrixSpec Hyper-Eye + Targan.
• Bees (1 CAAIN project): Nectar Technologies BeeTrack.

Total: 11 CAAIN-funded animal production projects + 4 existing processing-side + 1 US vendor deployment.

The processing cell is now multi-protein across animal proteins (SoraLINK dairy + P&P Optica meat + mode40 carcass cooling + MatrixSpec egg + Circulus Agtech manure-to-fertilizer). The corpus can cite 5+ protein categories with named Canadian deployments.

Talk-ready claims.

• “Canadian animal production AI covers the full value chain — production, processing, logistics.”
• “Dairy AI is the first value chain where CAAIN + existing units cover the full chain.”
• “Canadian processing AI is now multi-protein across 5+ protein categories.”

Anchor units. units/canadian-dairy-ai.md, units/canadian-meat-processing-ai.md, units/canadian-poultry-ai.md, units/canadian-beekeeping-ai.md, units/soralink-export-food-processing.md (existing).

Archetypes. 01 (vendor-sweep) — primary; 06 (regional-cluster-comparison) — secondary.

5. The MatrixSpec 7-billion-chick culling claim (headline for critical-lens)

MatrixSpec Hyper-Eye’s claim that 7 billion day-old male chicks are culled annually worldwide — and that pre-hatch hyperspectral imaging AI could eliminate this — is the most consequential animal-welfare AI claim in the corpus.

Dr. Michael Ngadi (McGill James McGill Professor, 25 years; born in Nigeria, Government of Canada scholarship 1991): “If hyperspectral imaging can be used to determine which eggs should be kept and which should be disposed of before they hatch, the impact will be extraordinary. The savings in time and money alone will be significant, as will the social and environmental improvements. We are very excited by the difference we will make when Hyper-Eye is eventually commercialized. This is a made-in-Canada solution to a global issue, and it wouldn’t have been possible without CAAIN.”

Talk-ready claims.

• “Hyper-Eye AI could eliminate the 7-billion-chick annual culling — a made-in-Canada solution to a global animal-welfare issue.”
• “The MatrixSpec pre-incubation sexing is structurally distinct from post-hatch chick sexing (Chick Pick + Targan WingScan).”
• “This is the most consequential animal-welfare AI claim in the corpus.”

Anchor unit. units/canadian-poultry-ai.md.

Archetypes. 05 (critical-lens) — primary; 01 (vendor-sweep) — secondary.

6. The RAII / Universal Broadband Fund tension (headline for adoption-diagnosis)

RAII is the federal AI SME funding layer that scales ($200M → $500M). The Universal Broadband Fund (UBF) is the rural connectivity layer that does not renew. Wire Report (industry source, 27 May 2026): “Ottawa will not renew its multi-billion dollar funding program designed to connect every household to high-speed internet by 2030.” ISED confirmed (5 June 2026) it will not raise the 50/10 Mbps minimum standard. Auditor General of Canada baseline (2021, still the reference): 90.9% national coverage, but only 42.9% on First Nations reserves and 59.5% in rural/remote.

For rural AI adoption to work, both layers must scale together. The CanNor Inuvik Tech Society project (“a region that only recently gained reliable, high-speed connectivity”) is a primary-source confirmation of this binding constraint.

Talk-ready claims.

• “Federal AI SME funding scales ($200M → $500M); rural connectivity funding does not renew (UBF May 2026).”
• “First Nations reserves coverage is 42.9% — less than half of the national 90.9% — and ISED won’t raise the 50/10 Mbps minimum standard.”
• “For rural AI adoption to work, RAII and UBF must scale together — they’re out of phase.”
• “The CanNor Inuvik Tech Society project’s ‘only recently gained reliable, high-speed connectivity’ language is a primary-source confirmation of the binding constraint.”

Anchor units. units/raii-canada-ai-adoption-programme.md, units/northern-canada-can-ai-2026.md, scouts/2026-07-canada-constraint-critical.md §2.

Archetypes. 03 (adoption-diagnosis) — primary; 02 (data-sovereignty) — secondary.

7. Indigenous-led agrifood AI deployment is now operational, not just framework (headline for critical-lens)

Pre-cycle, the G-010 corpus gap was real: Indigenous-led AI agrifood deployment was framework-only (CARE / IEEE 2890 / OCAP). Post-cycle:

• PolArctic / Sanikiluaq Nunavut mariculture AI — first AI model to treat Indigenous Knowledge and Western science as equals. Inuktitut + English. Pilot completed 2021.
• Salmon Vision BC Indigenous-led wild salmon AI — Frontiers in Marine Science 2023 (peer-reviewed). 12 species, 500,000+ frames, mAP 67.6%. Bear River + Kitwanga River (Gitanyow).
• Nunavut Economic Developers Association AI tool — first documented Indigenous-led AI deployment with federal funding (CanNor REGI-AI, $200K / 3 years).
• Inuvik Tech Society digital literacy + AI training — Gwich’in Settlement Region and Inuvialuit Settlement Region (CanNor IDEANorth).

G-313 (no federal Indigenous-led agrifood AI funder) remains real — none of these is funded through an Indigenous-led-agrifood-AI-specific programme. But G-010 (operational Indigenous-led AI agrifood deployment) is now substantively populated.

Talk-ready claims.

• “Indigenous-led AI deployment is now operational in Canada — PolArctic Sanikiluaq, Salmon Vision, Nunavut Economic Developers Association.”
• “PolArctic is the first AI model to treat Indigenous Knowledge and Western science as equals — a ‘two-eyed seeing’ deployment.”
• “Salmon Vision is peer-reviewed in Frontiers in Marine Science 2023 with named Gitanyow Fisheries Authority partnership.”
• “G-313 — no federal programme specifically funds Indigenous-led agrifood AI deployment — remains a structural gap.”
• “First documented Indigenous-led AI deployment with federal funding: Nunavut Economic Developers Association AI tool (CanNor REGI-AI).”

Anchor units. units/northern-canada-can-ai-2026.md, units/bc-horticulture-ai-cluster.md (Salmon Vision), units/canadian-aquaculture-ai.md.

Archetypes. 05 (critical-lens) — primary; 02 (data-sovereignty) — secondary.

8. The DFO Pacific salmon AI + 551-FTE reduction counterweight (headline for adoption-diagnosis)

DFO has three active AI pilot projects in the Pacific Salmon Strategy Initiative:

• Chumputer — deep-learning CNNs for salmon-scale age reading. 80,000+ scales/year.
• Computer-vision salmon migration counter — automated counting at Sproat and Stamp River fish ladders.
• Factoid Finder — NLP/LLM-style AI for watershed planning (Nicola Watershed IPSE).

DFO’s 2026-27 spending reduction: $54.47M (2026-27), $101.91M (2027-28), $193.82M (2028-29) / 551 FTE decrease by 2028-29. The AI pilots are positioned as the digital modernization counterweight to these cuts — AI is being deployed in part to substitute for reduced human capacity rather than to augment it.

Talk-ready claims.

• “DFO has three active Pacific salmon AI pilots — Chumputer, computer-vision migration counter, Factoid Finder.”
• “DFO faces a $54M / $102M / $194M spending reduction over 2026-29 / 551-FTE decrease; AI pilots are positioned as the digital modernization counterweight.”
• “Canadian federal AI is being deployed in part to substitute for reduced FTE capacity — a structural pattern worth naming.”
• “The 2026 DFO transition plan for 79 BC salmon farms is unresolved — operational future of Mowi + OnDeck + Catalera-dependent salmon-farm AI deployments uncertain.”

Anchor units. units/dfo-pacific-salmon-ai.md, units/canadian-aquaculture-ai.md, units/bc-horticulture-ai-cluster.md.

Archetypes. 03 (adoption-diagnosis) — primary; 06 (regional-cluster-comparison) — secondary.

9. The Saskatoon / Winnipeg / Calgary Prairie cluster (headline for regional-cluster-comparison)

The Prairie cluster has three sub-clusters, distinct from the BC horticulture cluster and the Ontario dairy/cereal cluster:

• Saskatoon grain AI — Super GeoAI Technology + VeriGrain + Raven Industries (OMNiPOWER) + Mojow Autonomous Solutions + University of Saskatchewan Crop Development Centre. The 100-year-old grain grading process is the legacy that the Saskatoon cluster challenges.
• Winnipeg/Steinbach/Vermilion protein AI — MacDon Industries ($9.26M, largest CAAIN-funded project) + mode40 (ISED Challenge → CAAIN → AAFC Lacombe → 10 commercial plants development path) + Lakeland College Precision Ranching.
• Calgary/Lethbridge/Olds digital ag — Ox+Plow + SmartGro Bioengineering + PIP International + Metabolomics + AltaML + AUAV Tech + Olds College Smart Farm.

MacDon Industries’ $9.26M total is the largest single CAAIN-funded project. CAAIN funding as the leveller (Weiping Zeng: “served as a lifeline for us at a time when our future was uncertain”).

The mode40 development path (ISED Challenge → CAAIN → AAFC Lacombe → commercial plants) is the full Canadian development pathway for federally-supported AI ventures.

Talk-ready claims.

• “Canada has three Prairie sub-clusters — Saskatoon grain AI, Winnipeg/Steinbach/Vermilion protein AI, Calgary/Lethbridge/Olds digital ag.”
• “MacDon Industries’ $9.26M is the largest single CAAIN-funded project.”
• “The 100-year-old grain grading process is the legacy that the Saskatoon cluster challenges — Super GeoAI + VeriGrain integration.”
• “mode40’s development path (ISED Challenge → CAAIN → AAFC Lacombe → 10 commercial plants) is the full Canadian development pathway for federally-supported AI ventures.”

Anchor units. units/prairie-grain-ai-cluster.md, units/canadian-grain-grading-ai.md.

Archetypes. 06 (regional-cluster-comparison) — primary; 03 (adoption-diagnosis) — secondary.

10. AIDA + BC horticulture cluster + Northern CanNor cycle (headline for regional-cluster-comparison)

Atlantic Canada (AIDA) — Atlantic Institute for Digital Agriculture at Dalhousie Faculty of Agriculture, Director Dr. Travis Esau, 13 Scientific Advisory Committee members + Neethirajan (14 total named scholars, correcting prior “11 Chair-holders” undercount), 5 research pillars (precision agriculture, robotics/automation/AI, data-driven management, human-computer interface, knowledge mobilization), Canadian Soil Data Portal (Heung CRC Tier II). NB Potato Industry Research Chair (Al-Mallahi). See units/aida-atlantic-digital-agriculture.md. Three critical-academic voices consolidated: Neethirajan (rural + Green AI; see units/neethirajan-dalhousie-ecosystem.md), Dara (greenhouse + cyber risk; see units/ai4food-guelph.md), Charlebois (data-deficit; see units/aal-dalhousie.md).

British Columbia — four sub-clusters: Surrey orchard (CropVue + OKSIR codling moth smart trap), Salmon Arm mushroom (4AG Robotics), Vancouver Island aquaculture training (Excel Career College AI for Aquaculture + BCSFA + First Nations), Vancouver biocontrol + salmon (Catalera Terramera spinout $8.8M Series A May 2025, Mowi Canada West Feed Centre, OnDeck Fisheries AI $1.5M PCAIS, Salmon Vision team).

Northern Canada — CanNor Feb 2026 cycle ($2.815M, 4 projects, YT/NWT/NU): Prosper NWT ($2.316M / 3 years AI entrepreneurship), Inuvik Tech Society ($200K / 2 years digital literacy + AI training), Nunavut Economic Developers Association ($200K / 3 years Inuit-led AI tool), DeltaVue Yukon ($100K / 1 year Arctic sensor platform). First documented Northern Canadian AI funding cycle with named recipients.

Every Canadian region now has a regional substrate unit.

Talk-ready claims.

• “AIDA is the substantive Atlantic anchor with 11 named Chair-holders — Director Esau at Dalhousie Faculty of Agriculture.”
• “The BC cluster has four sub-clusters — Surrey orchard, Salmon Arm mushroom, Vancouver Island aquaculture training, Vancouver biocontrol + salmon.”
• “Excel Career College’s AI for Aquaculture programme is the only documented Canadian aquaculture AI programme with explicit Indigenous-knowledge integration.”
• “Catalera (Terramera spinout) is the largest disclosed Vancouver AI-agtech venture round — $8.8M CAD Series A May 2025.”
• “The CanNor Feb 2026 cycle is the first documented Northern Canadian AI funding cycle — $2.816M across 4 projects.”

Anchor units. units/aida-atlantic-digital-agriculture.md, units/bc-horticulture-ai-cluster.md, units/northern-canada-can-ai-2026.md.

Archetypes. 06 (regional-cluster-comparison) — primary; 05 (critical-lens Indigenous) — secondary.

11. Substantive contested claims surfaced (carry through to talks)

These are corpus-level contested claims — the disagreement goes on the slide, per talks/README.md operating principle 6.

• C-278 (new): “CAAIN is the principal Canadian agrifood AI funder.” Counter: Scale AI + Digital + PIC + RAII together fund more in dollar terms; CAAIN’s distinction is automation/robotics/data-driven focus + physical validation (Pan-Canadian Smart Farm Network).
• C-280 (new): “Cluster programmes are independent.” Counter: at least 5 leads (Vivid Machines, Terramera, Agi3, Verge Ag, EarthDaily) appear in both Scale AI and Digital.
• C-282 (new): “PIC AI Programme is ongoing.” Counter: closed March 2026.
• C-286 (new): “Closed-loop circular-economy AI is not yet a Canadian pattern.” Counter: Circulus Agtech’s modular on-farm manure treatment system is the first documented Canadian circular-economy AI deployment.
• C-287 (new): “Atlantic Canada has no academic anchor for digital agriculture.” Counter: AIDA + 11 named Chair-holders.
• C-289 (new): “DFO has no AI deployment.” Counter: three active Pacific salmon AI pilots + DFO 2026-27 Departmental Plan explicit AI investment.
• C-290 (new): “Canada has no Indigenous-led agrifood AI.” Counter: PolArctic + Salmon Vision + Nunavut Economic Developers Association + Inuvik Tech Society.

12. Substantive gaps surfaced (carry through to talks as honesty, per talks/README.md principle 7)

• G-313: No federal or provincial programme specifically funds Indigenous-led agrifood AI deployment. Real gap.
• G-314: No federal farm-data-governance programme adopted. EMILI/CADI guidebooks are industry-led; UK Farm Data Principles equivalent recommended (CAPI/EMILI ISED submission) but not adopted.
• G-315: No CSA-equivalent safety standard for agrifood AI in Canada.
• G-316: No FCC AgExpert data-licensing public framework.
• G-317: RAII funding shares by RDA (other than PrairiesCan’s $33.8M) not publicly disclosed. Transparency gap.
• G-318: No federal funder channel covers post-consumer food waste AI specifically across CAAIN / Scale AI / Digital / PIC / RAII.
• G-319: No CAAIN project is Indigenous-led.
• G-320: No CAAIN project is in Newfoundland and Labrador, PEI, Yukon, NWT, or Nunavut. CAAIN’s geographic coverage is 8 of 13 provinces/territories.
• G-321: Operational scale of CAAIN projects not publicly disclosed for many projects (4AG Robotics, Nectar, Circulus, A.U.G. Signals, Grain Discovery).
• G-322: AIDA’s specific industry partnerships and deployment outcomes not publicly detailed.
• G-323: DFO’s three Pacific salmon AI pilots are pilot-stage; operational scale not tracked.
• G-324: PolArctic Sanikiluaq operational status (production mariculture vs. pilot-completed) unclear.
• G-325: OnDeck Fisheries AI’s $1.5M PCAIS-funded project is positioned as building infrastructure, not deployed.
• G-326: Mowi Canada West Feed Centre operational AI deployment scale is industry-framing only. 2026 DFO transition plan for 79 BC salmon farms unresolved.
• G-327: Salmon Vision’s stated expansion to Koeye and KwaKwa rivers is forward-looking; needs verification.
• G-328: Canadian Soil Data Portal provincial coverage and accessibility tiers not publicly documented.

13. Archetype-specific upgrade notes

Archetype 02 (data-sovereignty)

Add to segment 2 (“what data is being collected”): Canadian aquaculture funding-stack pattern. The same CV-on-fish capability looks structurally different depending on the funding regime. Anchor unit: units/canadian-aquaculture-ai.md. Finding 3 above.

Add to segment 3 (“the contract layer”): FCC AgExpert 25,000 users / 6.5M acres — data-licensing terms are private contract terms. G-316. The absence of a public FCC data-licensing framework is structurally distinct from Bayer/Deere/AGCO vendor contracts.

Add to segment 4 (“the cooperative alternative”): The CAAIN founder-pattern observation (immigrant / multi-generation ag family × technical training × federal funding as the leveller) is the substantive Canadian producer-led alternative posture. Anchor unit: units/caain-portfolio-canada.md.

Archetype 03 (adoption-diagnosis)

Upgrade segment 1 (“the headline”): The 1.8% / 12.2% gap is still the headline, but now contextualised by the five-funder substrate (Finding 1) and the RAII / UBF tension (Finding 6).

Add segment 2.5 (“the funder substrate”): CAAIN + Scale AI + Digital + PIC + RAII. Federal AI SME funding is scaling; rural connectivity is contracting.

Add segment 3.5 (“DFO Pacific salmon AI”): Digital modernization counterweight to FTE reduction. Finding 8.

Upgrade segment 4 (“policy levers”): The binding-constraint dynamic (RAII scales; UBF doesn’t renew; First Nations reserves coverage 42.9%) is the substantive policy lever.

Archetype 04 (cooperative-alternative)

Add Canadian anchor units: units/canadian-dairy-ai.md (CATTLEytics equipment-agnostic ingestion as the producer-led alternative), units/caain-portfolio-canada.md (CAAIN as federal-third-party-delivery alternative to venture capital), units/canadian-grain-grading-ai.md (Super GeoAI 100-year-old grain grading challenge).

Note: Canadian cooperatives as AI actors is structurally thin in the corpus. The CAAIN founder-pattern (federal funding as the leveller) is the closest Canadian analog to Mondragón’s institutional-federation-anchor. Worth tracking.

Archetype 05 (critical-lens Indigenous sovereignty)

Major upgrade. G-010 (Indigenous-led AI agrifood deployment) is now substantively populated:

• PolArctic / Sanikiluaq — Inuit-led mariculture AI; first AI to treat Indigenous Knowledge and Western science as equals. Findings 7 + 10.
• Salmon Vision BC — Indigenous-led wild salmon AI; Frontiers in Marine Science 2023 peer-reviewed; Gitanyow Fisheries Authority partnership.
• Nunavut Economic Developers Association — first documented Indigenous-led AI deployment with federal funding.

Add segment on funding gap (G-313): No federal or provincial programme specifically funds Indigenous-led agrifood AI deployment. The Indigenous-led AI deployments documented are all funded through non-Indigenous-specific programmes (BCSRIF + DFO + WWF-Arctic + Qikiqtaaluk + CanNor REGI).

Archetype 06 (regional-cluster-comparison)

Add Canada cluster pattern observation: Federally orchestrated + cluster-channeled + regionally delivered. Distinct from the EU’s DG-distributed substrate and the US private-venture substrate.

Add cross-cluster funding pattern: At least 5 leads in both Scale AI and Digital. Same companies, multiple federal funding streams. Finding 2.

Add regional sub-clusters: Saskatoon grain AI; Winnipeg/Steinbach/Vermilion protein AI; Calgary/Lethbridge/Olds digital ag (Prairies); AIDA (Atlantic); Surrey/Salmon Arm/Vancouver Island/Vancouver (BC); YT/NWT/NU (North); Halifax NS (Atlantic aquaculture); Vancouver (Pacific fisheries).

Add funding-stack pattern: Venture + PCAIS + corporate + cluster + NGO + Indigenous-led — six distinct Canadian aquaculture AI funding regimes producing structurally different data-governance postures. Finding 3.

14. Quick-reference unit index (for talk-building)

Funder/convenor substrate (5 units):

• units/caain-portfolio-canada.md — 35+ projects, $19M+ CAAIN contribution.
• units/scale-ai-agriculture-canada.md — 13 Agriculture & Mining projects.
• units/digital-supercluster-agriculture-canada.md — 20+ Natural Resources & Agriculture projects.
• units/protein-industries-canada-ai-programme.md — $30M PCAIS-delivered, ended March 2026.
• units/raii-canada-ai-adoption-programme.md — $200M → $500M, 7 RDAs.

Value-chain matrix (7 units):

• units/canadian-dairy-ai.md — full value chain (production + processing + logistics).
• units/canadian-meat-processing-ai.md — hyperspectral + carcass cooling + circular-economy.
• units/canadian-poultry-ai.md — chick sexing + disease mapping + pre-incubation sexing.
• units/canadian-beekeeping-ai.md — first documented Canadian AI beekeeping deployment.
• units/canadian-mushroom-ai.md — 4AG Robotics vision-guided mushroom AI.
• units/canadian-orchard-ai.md — codling moth smart traps + tree-fruitlet CV.
• units/canadian-grain-grading-ai.md — 100-year-old grain grading legacy challenged.

Regional substrate (6 units):

• units/aida-atlantic-digital-agriculture.md — AIDA + 13 SAC + Neethirajan.
• units/uog-bean-gpt-najafabadi.md — BeanGPT + Najafabadi lab (inputs-cell).
• units/programmatic-breeding-ai.md — conceptual umbrella.
• units/prairie-grain-ai-cluster.md — three Prairie sub-clusters.
• units/bc-horticulture-ai-cluster.md — four BC sub-clusters.
• units/northern-canada-can-ai-2026.md — CanNor Feb 2026 cycle + PolArctic + Salmon Vision.
• units/dfo-pacific-salmon-ai.md — three DFO Pacific salmon AI pilots.
• units/canadian-aquaculture-ai.md — Pacific + Atlantic vendor ecosystem.

15. What to read next

For a 45-min Canadian talk:

1. units/fcc-canada-ai-adoption.md (existing FCC cycle scan anchor) — the 1.8% / 12.2% headline.
2. units/caain-portfolio-canada.md — the federal substrate.
3. units/raii-canada-ai-adoption-programme.md — RAII / UBF tension.
4. units/canadian-dairy-ai.md — full value-chain exemplar.
5. units/canadian-poultry-ai.md — animal-welfare claim.
6. units/northern-canada-can-ai-2026.md — Indigenous-led AI.

For a 60-min critical-lens talk:

1. units/northern-canada-can-ai-2026.md — PolArctic + Salmon Vision + Nunavut Economic Developers Association.
2. units/canadian-aquaculture-ai.md — funding-stack pattern.
3. scouts/2026-07-canada-constraint-critical.md §10 — Indigenous data sovereignty + farm-data-governance gaps.
4. units/raii-canada-ai-adoption-programme.md — RAII scaling + UBF non-renewal.
5. Existing units/canadian-retail-ai-pattern.md (existing) — vendor-mediated consumer data governance.

For a 60-min regional-cluster-comparison talk:

1. units/caain-portfolio-canada.md — Canadian funder substrate.
2. units/prairie-grain-ai-cluster.md — Prairie sub-clusters.
3. units/bc-horticulture-ai-cluster.md — BC sub-clusters.
4. units/aida-atlantic-digital-agriculture.md — Atlantic anchor.
5. units/northern-canada-can-ai-2026.md — Northern.
6. units/dfo-pacific-salmon-ai.md — federal regulator AI.
7. units/canadian-aquaculture-ai.md — Pacific + Atlantic vendor ecosystem.

Canada academic-research substrate cycle (July 2026 cycle)

Anchor units (new in this cycle):

• units/aida-atlantic-digital-agriculture.md — Atlantic academic+convening substrate (13 SAC + Neethirajan; 5 research pillars; Canadian Soil Data Portal).
• units/ai4food-guelph.md — Ontario academic anchor (Rozita Dara; 4 focus areas; CCMPS Research Impact Leadership Chair).
• units/aal-dalhousie.md — critical-voice academic-policy lab (Charlebois; Canada’s Food Price Report; data-deficit frame).
• units/mila-quebec-agrifood.md — Quebec academic AI hub (Mila + IVADO + Innovasea; DISA Rwandan-deployment clarifier; corrects G-029 → G-331).
• units/canada-academic-research-funding-stack.md — NSERC + Mitacs + CFI + CFREF + FRQ + AAFC academic-research funding tier.
• units/uog-bean-gpt-najafabadi.md — BeanGPT + Najafabadi lab (first named AI4Food deployment-tier product; 9 named partners; 6 named AI projects).
• units/programmatic-breeding-ai.md — conceptual/methodological unit (programmatic + predictive + AI-assisted selection pipelines).
• scans/2026-07-canada-academic-research-substrate.md — consolidating scan.

Substantive findings:

• Five-anchor academic substrate: AIDA (Atlantic) / AI4Food (Ontario) / AAL (Dalhousie critical-voice) / Mila + IVADO (Quebec PCAIS) / Olds College (Prairie applied-research). Distinct from the federal-cluster deployment substrate (CAAIN / Scale AI / Digital / PIC / RAII).
• Three critical-academic voices consolidated: Neethirajan (rural + Green AI) / Dara (greenhouse + cyber risk) / Charlebois (data-deficit). Three frames, three institutions.
• Cluster-vs-research funding-stack distinction: cluster funding (TRL 7+) and academic-research funding (TRL 3–6) are complementary, not duplicative. No dedicated federal academic-research agrifood AI funding channel. (G-332)
• Federal-AIDA / academic-AIDA acronym collision flagged for talks clarity.
• Mila DISA reframed: Canadian-research-pipelined Rwandan-deployment, not Canadian farm deployment. G-029 superseded by G-331.
• BeanGPT surfaces as first named AI4Food deployment-tier product. Canadian seed-breeding AI is academic-research-led + provincial-commodity-pipelined (BeanGPT + Ontario Bean Growers + Hensall Co-Op + Sprague Foods), distinct from the Brazilian seed-AI cluster-with-three-structures pattern.
• Programmatic-breeding-AI is the conceptual umbrella covering CABI 2026’s “programmatic breeding goals” + 2026 “predictive breeding” (Springer; Garcia-Oliveira) + “AI-assisted selection pipelines” (Computers Electronics in Agriculture sugarcane review) + 2024-2026 generative-AI breeding platforms (BeanGPT-class). BeanGPT = generative-AI query-decision layer of a full programmatic-breeding pipeline.

AI plant breeding global scan (July 2026 cycle)

Anchor units (8 new in this cycle):

• units/plant-breeding-ai-methodology.md — substantive 2026 academic methodology stack (phenomics + genomics + multi-omics + GS + AI/ML + generative AI + CRISPR-AI).
• units/longping-yuan-caas-china-seed-ai.md — Chinese state-orchestrated cluster; CAAS + Syngenta Group + BGI/MGI + smart-breeding national plan.
• units/cgiar-eib-global-south-plant-breeding.md — CGIAR Breeding for Tomorrow + EiB; 12 priority crops + 700K+ germplasm + Gates/FFAR/USAID substrate.
• units/limagrain-kws-ragt-eu-private-plant-breeding.md — EU private-breeding cluster; cooperative + family-controlled + public-academic.
• units/bayer-syngenta-corteva-multinational-pipelines.md — Bayer + Syngenta Group + Corteva + BASF + ETC/GRAIN 2025 concentration observation.
• units/usda-ars-iowa-state-aiira-us-land-grant.md — US land-grant cluster; USDA-NIFA AI Institutes + USDA-ARS + Heritable Agriculture (Google X).
• units/indigenous-seed-sovereignty-ai-breeding.md — cross-cutting critical-voice; ITPGRFA + CARE + DSI + IPES-Food 2026.
• units/ai-breeding-genetic-diversity-counter-narrative.md — cross-cutting counter-narrative; concentration + narrowing + reproducibility.
• scans/2026-07-ai-plant-breeding-global.md — consolidating scan.

Substantive findings (6 substantively distinct regional cluster shapes globally):

1. Chinese state-orchestrated (CAAS + Syngenta Group + BGI/MGI + smart-breeding national plan; Yuan Longping legacy; Sinochem/ChemChina ownership; 2021 amended Seed Law IP-strengthening).
2. CGIAR public-platform for Global South (12 priority crops including cassava/yam/sweetpotato/cowpea/millet/sorghum; 700K+ germplasm in ITPGRFA trust; ICRISAT × CIMMYT AI-driven predictive breeding initiative; Gates/FFAR/USAID funding substrate; 15-30% genetic-gain uplift commitment).
3. EU private-cluster (Limagrain cooperative via InVivo; KWS/Enza Zaden/Rijk Zwaan family-controlled; Wageningen + INRAE public-academic; NGT Regulation 2026 substantive EU policy frame).
4. Multinational-corporate-pipelined (Bayer ~€6-11bn R&D pipeline 2024-2030+; Syngenta Group Chinese state-owned; Corteva Granular on-farm AI; BASF xarvio; ETC/GRAIN 2025 “Four firms control 56% of global seed market”).
5. US land-grant (USDA-NIFA + USDA-ARS + land-grant + extension + corporate-startup: AIIRA at Iowa State; Heritable Agriculture from Google X; Biographica + BASF; Texas A&M wheat-breeding pipeline genomic prediction).
6. Canadian provincial-commodity-pipelined (BeanGPT + Ontario Bean Growers + ARIO + OMAFA + Hensall Co-Op + Sprague Foods + Agilent + NSERC; substantively distinct from multinational-corporate-pipelined or state-orchestrated clusters).

Substantive cross-cluster bridging observations:

• Syngenta Group bridges Chinese state-orchestrated + multinational-corporate (substantively global-bridging entity).
• Heritable Agriculture bridges US land-grant + corporate-startup + multinational-corporate (through Syngenta Vegetables + Heritable Ag collaboration 2026).
• CGIAR Multilateral System bridges multilateral-system + public-domain + farmer-led seed networks.
• BeanGPT bridges Canadian provincial-commodity + CGIAR public-platform in concept (both substantively public-platform-adjacent vs. corporate-multinational).

Substantive 2026 cross-cutting tension observation: speed-of-cycle × diversity-preservation × reproducibility × concentration × DSI governance × CARE Principles for Indigenous. Per Varshney 2026 (Rewiring diversity, physiology, and practice): “how to expand genetic diversity and enhance stress resilience while maintaining yield stability, quality.”

Substantive 2024-2026 DSI multilateral-mechanism negotiation substantively determines the substantive governance frame for AI plant breeding; 2026-2030 deployment will determine whether AI plant-breeding pipelines substantively apply multilateral-system governance, bilateral-ABS, sovereign-state, or open-access.

8 substantive AI-methodology-anchor cluster shapes substantively across the global scan:

• RAG over literature (BeanGPT).
• Knowledge-graph LLM (Xie 2026 Plant Communications).
• Digital-twin (AIIRA + Heritable Agriculture).
• Generative-AI Q&A (Crop GraphRAG 2025).
• CRISPR-AI integration (Huang/Kamran 2026).
• Genomic prediction in applied European wheat (Thomsen 2026).
• AI-driven predictive breeding at scale (CGIAR Breeding for Tomorrow).
• Industry-vertical AI deployment (Bayer/Syngenta/Corteva).