Michael Carolan — Colorado State University sociologist of digital agriculture, substantive US critical voice

NA-US (Colorado State University; research reach international)

Content

Michael Carolan is Professor of Sociology and Co-Director of the Food Systems Institute at Colorado State University. With 300+ peer-reviewed publications, he is the most substantive US-based sociological critical voice on digital agriculture. Per memory hygiene (substantive analytical contribution over citation volume): Carolan’s work is included because he engages substantively with the labour, recognition, and disability dimensions of agrifood AI deployment — dimensions that mainstream agrifood AI reports (USDA AI Strategy, the AI Institutes’ communications, vendor marketing) systematically avoid.

Key US-based works (selected)

“Who and what gets recognized in digital agriculture: Agriculture 4.0 at the intersectionality of (Dis)Ableism, labor, and recognition” — Agriculture and Human Values, 2024. Carolan’s most substantive recent US-engaged work. Engages with how digital agriculture’s “recognition” structures (what counts, what is measured, whose work is visible) systematically exclude disabled workers, farmworkers with limited formal documentation, and care work. The (Dis)Ableism framing is structurally distinctive — most agrifood AI analytical work does not engage with disability.

“Digital agriculture killjoy: Happy objects and cruel quests” — Sociologia Ruralis, 2023. A substantive critique of the affective politics of digital agriculture — what Carolan calls the “happy objects” framing of AI / robotics / data tools, and the “cruel quests” that follow when these framings meet labour realities on farms.

“The perilous promise of productivity: Affective politics” — New Media & Society, 2023. Engages with the productivity framing that dominates agrifood AI discourse. The “perilous promise” framing is substantive — it names the political economy of productivity claims in digital agriculture.

Why Carolan matters for the field guide

The dominant voices in US agrifood AI discourse are:

• Vendors (Bayer, John Deere, Indigo, AGCO, etc.) — productivity-and-efficiency framings
• Federal agencies (USDA AI Strategy) — workforce-readiness-and-ethics framings
• Academic research institutes (AgAID, AIIRA, AIFARMS, etc.) — adopt-adapt-amplify / digital-twin / autonomous-farming framings
• Land-grant deployment (Cooperative Extension) — people-trust-coordination framings

What isn’t there: substantive sociological / labour-conditions / disability-and-recognition critique from US-based researchers. Carolan fills that gap. His work does not duplicate what Tzachor (Cambridge / Reichman; UK / Israel) or Bronson (Ottawa; Canada) do — Tzachor is more on responsible-AI risk frameworks; Bronson is more on data-governance. Carolan is more on the social relations of agrifood AI deployment — who is recognised, who is excluded, what the productivity promise actually delivers.

Institutional position

• Colorado State University — Fort Collins, CO. Land-grant university (one of the 1862 institutions).
• Food Systems Institute — Co-Director. The Institute is CSU’s interdisciplinary food systems research hub.
• AI-LEAF adjacency — CSU is a partner in the AI-LEAF Institute (see units/ai-leaf-penn-state-institute.md). Carolan’s Food Systems Institute is institutionally adjacent to AI-LEAF’s CSU partner work but analytically distinct — Carolan’s work is critical-sociological, AI-LEAF’s CSU work is technical / applied.

What this unit is doing in the taxonomy

Anchors the US academic critical-sociological analysis of agrifood AI cell. Distinct from:

• Tzachor et al. (2022) Responsible artificial intelligence in agriculture — Cambridge / Reichman (UK / Israel); responsible-AI risk framework focus. Cited 266 times; most-cited substantive analytical work on responsible ag AI globally, but not US-anchored.
• Bronson (University of Ottawa) — substantive US-adjacent but Canadian-anchored; data-governance focus. See quotes/_index.md for adjacent analytical voices.
• Alexander et al. (2024) Who is responsible for “responsible AI”? — Cited 98 times; trust-in-AI focus. US authorship unconfirmed; potentially Canadian / international collaboration.
• Indigenous Data Sovereignty (units/indigenous-data-sovereignty.md) — IDSov framework; structurally parallel but distinct (Indigenous-sovereignty rather than labour-and-recognition).
• USDA AI Strategy (units/usda-fy25-26-ai-strategy.md) — institutional federal framing; does not substantively engage labour / disability / recognition critique.

Why it matters for talks

• Carolan is the canonical US-based critical-voice researcher for agrifood AI sociology. Talks that need substantive critical analytical content (academic, policy, or critical-civil-society audiences) should reference Carolan’s framing.
• The (Dis)Ableism framing is structurally distinctive. Most agrifood AI critique focuses on smallholder / Indigenous / data-sovereignty concerns; Carolan adds labour and disability dimensions that are genuinely distinct.
• Carolan’s CSU position connects him institutionally to the AI-LEAF cluster; the proximity is worth knowing for talks that pair technical and critical framings.

Critical context

• Carolan is not a researcher who deploys AI. He is a critical analyst. Treating his work as a counterweight to vendor / institute framings is appropriate; treating it as a substitute for substantive technical work is not.
• The 300+ publications indicator includes a wide range of topics — food systems generally, not just digital agriculture. The digital agriculture subset is substantive but is a portion of broader work.
• Carolan’s framing of “the perilous promise of productivity” is a political-economy critique, not a rejection of agrifood AI per se. Worth knowing because his work can be misread as anti-technology when it is anti-uncritical-adoption.
• Carolan’s positionality: he is US-based but his work engages substantively with global food systems. The “US-based” classification in this unit is institutional-location; his analytical reach is broader.
• The unit does not count Carolan’s work toward the v4 taxonomy’s actor-type: academic quota in the same way as the AI Institutes — Carolan’s claim-type: claim rather than example reflects that his work is analytical rather than deployment-describing.