Jason Miklian, New ResearchNew publications, working papers, and concepts from Jason Miklian, Research Professor at the University of Oslo. Business and peace, polycrisis, fragile states, AI governance.https://miklian.org/2026-10-05T00:00:00ZJason Miklianhttps://miklian.org/jason.miklian@globe.uio.noCreative Commons Attribution 4.0 International (CC BY 4.0)Business, Politics and Corporate Power: a new research area pagehttps://miklian.org/business-and-politics/2026-10-05T00:00:00Z2026-10-05T00:00:00ZA new research area page gathering seven papers on how firms bargain with governments and communities: data center moratoria in 348 US counties, corporate silence under polarization, charter cities and network states, platforms in hybrid regimes, Santa Anna and foreign nationals in Mexico, and ten of Marx's hypotheses tested against twenty-first-century data.Polycrisis (concept)https://miklian.org/concepts/polycrisis/2026-10-05T00:00:00Z2026-10-05T00:00:00ZSeveral crises running at once and feeding each other, so that a response built for one worsens another. The definition and the evidence behind it: the 2021 Harvard Business Review crisis playbook, the Beirut study of 34 SME owners, the Cambridge Element on ethical leadership in conflict and crisis, and the Crisis casebook with its Force Majeure simulation.When the Models Become Kingmakers: Assessing Textual Similarity and Variation in AI LLM Academic Reference Selectionhttps://miklian.org/papers/when-the-models-become-kingmakers-ai-llm-academic-reference-selection/2026-10-05T00:00:00Z2026-10-05T00:00:00ZSix leading LLMs (GPT, Claude, Gemini, Grok, DeepSeek and Qwen) chose ten references from the same thirty candidates for 396 papers across six fields. Model pairs agreed on 6.45 of ten selections, 36.7 percent of candidates were chosen by none of the six, textual similarity to the manuscript predicted selection better than any other measure, references from other disciplines were chosen 14.1 points less often, and replacing a paper's shared vocabulary lowered its selection rate by 4.8 points. Displayed affiliation and gender-coded names made little difference in balanced pools. Early discussion draft, 5 October 2026, submitted to arXiv and SSRN.Data Center Resistance in the United States: Local Refusal and the Politics of Moratoriahttps://miklian.org/papers/data-center-resistance-in-the-united-states-local-refusal-and-the-politics-of-moratoria/2026-10-05T00:00:00Z2026-10-02T00:00:00ZBy August 2026, local governments in 348 United States counties had adopted 457 moratoria and other restrictions on data centers, 388 of them in 2026. Matched to an inventory of 2,165 projects, counties adopted a first restriction most often in the six months after a project announcement, 193 of the 348 adopting counties had no recorded project, restrictions spread between counties where the same opposition group was recorded, and projects in counties that restricted early were 9.46 percentage points less likely to have started construction. Early discussion draft, 2 October 2026, submitted to arXiv and SSRN.Solar extraversion: incumbent interests, foreign capital and the making of Morocco's megaproject strategyhttps://miklian.org/papers/solar-extraversion-incumbent-interests-foreign-capital-moroccos-megaproject-strategy/2026-09-04T00:00:00Z2026-09-04T00:00:00ZMorocco legalized decentralized solar generation and prioritized the Noor Ouarzazate megaproject anyway. Incumbent producers and distributors blocked reform inside a fragmented state lacking adjudication mechanisms, while megaprojects returned concessional finance and geopolitical standing mobilized for domestic ends. The scale at which a project is evaluated shapes what counts as success. With Moritz Schmoll and Kristian Hoelscher, in Globalizations. Open access, CC BY-NC-ND 4.0.How Artificial Intelligence LLM Engines Shape the Global Conflict Information Environmenthttps://miklian.org/papers/how-ai-llm-engines-shape-the-global-conflict-information-environment/2026-07-16T00:00:00Z2026-07-16T00:00:00ZFive leading AI answer engines were asked 13 questions about 28 conflicts, three times each: 5,460 answers scored against documented evidence ranges. The engines are least accurate on the least-covered conflicts, miss settled facts precisely where coverage is thin, and the thinnest records are the most structurally exposed to Generative Engine Optimization, with state organs and belligerents among the heaviest optimizers. All leading AI LLMs amplify the world's neglect of most conflicts: the failure mode is amplification rather than hallucination."That's AI Slop, You Bot!": Studying Accusations, Evidence, and Credibility in Online Discourse Towards LLM-Generated Commentshttps://miklian.org/papers/thats-ai-slop-you-bot-studying-accusations-evidence-and-credibility-in-online/2026-06-10T00:00:00Z2026-06-10T00:00:00ZFirst large-scale observational study of AI slop accusations. Across 25 million Hacker News and Reddit comments (2023-2026), pejorative AI accusations rose more than tenfold and consolidated around the slop frame, yet a matched-control test shows the prose features that statistically distinguish AI text do not predict which human text gets accused. Accusation works as social gatekeeping, not detection. With John E. Katsos.Right Diagnosis, Wrong Prescriptions? Testing Ten of Karl Marx's Hypotheses Against Twenty-First-Century Datahttps://miklian.org/papers/right-diagnosis-wrong-prescriptions-testing-marx-hypotheses-21st-century-data/2026-04-14T00:00:00Z2026-04-14T00:00:00ZTen Marxist hypotheses tested on 21st-century panel data covering 183 countries. Economic diagnostics hold up strongly (mean .516); political predictions score far weaker (.300). Institutional design, not historical inevitability, mediates whether economic exploitation produces political consequences.A Problem of the Present: What Artificial Intelligence Tools Can (and Can't) Deliver for Business Under Crisis and Political Instabilityhttps://miklian.org/papers/problem-of-the-present-ai-tools-business-crisis-political-instability/2026-04-14T00:00:00Z2026-04-13T00:00:00ZAI tools synthesize the recent past and extrapolate probabilistic futures, but struggle with three epistemic demands instability places on decision-makers. Algorithmic mediation renders the most consequential stakeholders structurally undetectable. Local knowledge networks function as an insurance policy against catastrophic misreading at a fraction of the cost of algorithmic failure.What is an Academic Article For? The Production of Scholarly Work and its Meaning in an Agentic AI Worldhttps://miklian.org/papers/what-is-an-academic-article-for-scholarly-work-agentic-ai/2026-04-04T00:00:00Z2026-03-15T00:00:00ZSeven-part typology of scholarly activities mapped against AI capability, with empirical analysis of 2,750 articles showing AI linguistic markers doubling across social science journals.Digital Governance in Hybrid Regimes: How Global Social Media Platforms Navigate the Space Between Democracy and Authoritarianismhttps://miklian.org/papers/digital-governance-in-hybrid-regimes-social-media-platforms-democracy-authoritarianism/2026-03-01T00:00:00Z2026-03-01T00:00:00ZDevelops a "worst of both worlds" framework showing how hybrid regimes deploy authoritarian digital control through formally democratic legal mechanisms.A New Digital Divide? Coder Worldviews, the 'Slop Economy,' and Democracy in the Age of AIhttps://miklian.org/papers/a-new-digital-divide-coder-worldviews-the-slop-economy-and-democracy-in-the-age/2026-02-15T00:00:00Z2026-02-15T00:00:00ZIntroduces the slop economy and coder worldviews as analytical concepts for the political economy of AI in the Global South.Business Survival Strategies in a Polycrisis: SME Experiences from Beirut, Lebanonhttps://miklian.org/papers/business-survival-strategies-in-a-polycrisis-sme-experiences-from-beirut-lebanon/2025-09-01T00:00:00Z2025-09-01T00:00:00ZBeirut SMEs surviving polycrisis employed four distinctive strategies: hyper-localization, rapid revenue diversification, fierce team protection, and active community partnerships.Unlocking the Peace Premium: An Evidence-Based Review of the Potentials and Pitfalls of Private Sector Finance for Peacebuildinghttps://miklian.org/papers/unlocking-the-peace-premium-an-evidence-based-review-of-the-potentials-and/2025-06-01T00:00:00Z2025-06-01T00:00:00ZDefines and operationalizes the peace premium, the economic returns generated when private sector investment actively contributes to conflict transformation.