AI Race Scorecard Event Log — May 2026

Mark Howard

— 20 min read
Share

This event log is the detailed companion to the Scorecard. The Scorecard gives the month-end reading. The log shows the individual events behind that reading.

The question behind each item is simple: does this event tend to concentrate AI-related power, or does it help disperse capability, accountability, or public control?

The scores are directional. They are not meant to look more precise than the evidence allows.

How to read the scores

  • Concentration +3: A major concentration event. Durable, scalable, and institutionally important.
  • Concentration +2: A significant concentration event.
  • Concentration +1: A mild or emerging concentration event.
  • Mixed 0: Important, but balanced or still too uncertain to call.
  • Dispersion +1: A mild or emerging dispersion event.
  • Dispersion +2: A significant dispersion event.
  • Dispersion +3: A major dispersion event. Durable, scalable, and institutionally important.

A concentration score does not mean an event is simply “bad.” A dispersion score does not mean an event is simply “good.” The score asks a narrower question: who gains power, capacity, leverage, or voice as AI spreads?

May 2026 scored events

1. Colorado AI-law rewrite passes and is signed by Governor Polis

  • Source and date: Colorado Sun, May 12, 2026; Littler, May 15, 2026; CBS Colorado, May 15, 2026
  • Where this happened: Colorado / state AI regulation / automated decision-making
  • Sources:
  • Initial reading: Concentration +2
  • Scorecard reading: Concentration +2
  • Confidence in this reading: High
  • How stable this reading is: Medium-high
  • Last reviewed: 2026-05-18
  • Next planned review: 2026-08-15
  • Themes: regulation, state-capacity, democratic-accountability, labor, data, public-sector, market-structure
  • What happened: Colorado’s compromise rewrite of its AI law narrowed the earlier disclosure regime and was signed by Governor Jared Polis on May 14, 2026. SB 26-189 repeals and replaces the 2024 Colorado AI Act before implementation, shifts the framework toward automated decision-making technology affecting consequential decisions, delays the effective date to January 1, 2027, and emphasizes notice, adverse-action/human-review procedures, and record retention rather than broader high-risk AI governance.
  • Why this belongs in the scorecard: The bill has now both passed and been signed into law. The law changed the accountability structure for consequential AI decisions, and the move from systemic transparency toward notice-and-appeal is institutionally meaningful.
  • Why this points toward concentration: Colorado moved from the country’s most ambitious state AI governance framework toward a narrower, lighter-touch automated-decision law after sustained pressure from business and technology interests. Reducing public/systemic disclosure makes it harder for outside actors, journalists, advocates, and affected communities to evaluate institutional AI use. Accountability is individualized rather than civic or structural.
  • Why the story is not one-sided: Appeal rights, consumer notice, human-review procedures, and record-retention obligations may still create some accountability, especially if implementation and enforcement are strong.
  • What could change this reading: Strong implementing rules, public reporting, active enforcement, clear appeal procedures, or civil-society use of appeal data could weaken the concentration score. Weak enforcement, broad carveouts, minimal public reporting, further delays, or similar rollback pressure in other states would strengthen it.

2. Colorado data-center bills fail

  • Source and date: Colorado Newsline / Kiowa County Press, May 12, 2026
  • Where this happened: Colorado / data centers / infrastructure regulation
  • Source: https://kiowacountypress.net/content/both-colorado-data-center-bills-rejected-final-days-2026-legislative-session
  • Initial reading: Concentration +2
  • Scorecard reading: Concentration +2
  • Confidence in this reading: Medium-high
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-13
  • Next planned review: 2026-08-15
  • Themes: compute, regulation, state-capacity, democratic-accountability, market-structure
  • What happened: Both Colorado data-center bills failed in the final days of the 2026 legislative session after extended negotiations, lobbying, hearings, and attempted compromise.
  • Why this belongs in the scorecard: The failure of statewide data-center governance is a clear institutional outcome, not merely an announcement.
  • Why this points toward concentration: AI compute infrastructure continues to expand while state-level governance fails to mature. This is a strong example of physical AI infrastructure outrunning public regulatory capacity.
  • Why the story is not one-sided: Local governments may still develop siting, utility, disclosure, or community-benefit rules.
  • What could change this reading: Local moratoria, utility disclosure rules, community-benefit agreements, renewed state legislation, or strong municipal governance would weaken the concentration score. Rapid project approvals without transparency would strengthen it.

3. Denver City Council approves one-year data-center moratorium

  • Source and date: Axios Denver, May 19, 2026; Government Technology / Denver Post, May 19, 2026; Denver City Council Legistar file, May 18, 2026
  • Where this happened: Denver, Colorado / local government / compute infrastructure / land use / environmental justice
  • Sources:
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: High
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-19
  • Next planned review: 2026-08-15
  • Themes: compute, local-government, regulation, civic-infrastructure, democratic-accountability, public-infrastructure
  • What happened: Denver City Council unanimously approved a one-year moratorium on new data-center development and construction. The moratorium pauses new qualifying data-center projects while the city studies potential rules for energy use, water use, diesel backup generation, neighborhood impacts, environmental justice, and other infrastructure concerns. The pause does not affect data centers already operating or already permitted, including the CoreSite project at 4900 Race Street.
  • Why this belongs in the scorecard: This item was previously on the Watchlist because Council action had not yet been finalized. The Council vote is now a concrete governance action that changes Denver's permitting path for new data-center projects for one year.
  • Why this points toward dispersion: Denver is using local land-use authority to slow the default compute-infrastructure buildout path long enough to create public rules. The significance is procedural and institutional: resident concern about power demand, water use, diesel generators, air pollution, and neighborhood burden has been converted into a formal governance pause. This adds civic-infrastructure friction to a developer- and utility-led deployment path.
  • Concentration risks / limits: The moratorium excludes already operating or permitted facilities, so it does not reverse projects already far along in the pipeline. It could also expire without durable rules, shift development pressure to less-regulated nearby jurisdictions, or be weakened through carveouts.
  • What could change this reading: The score would strengthen if Denver adopts enforceable standards for power use, water use, diesel backup generation, public disclosure, cumulative neighborhood burden, ratepayer protection, or community-benefit requirements. It would weaken if the moratorium expires without meaningful rules, is narrowed by exemptions, or mainly displaces projects without changing governance capacity.

4. Utah launches statewide AI Workforce Credential

  • Source and date: Utah System of Higher Education, May 1, 2026
  • Where this happened: State higher education / workforce credentialing
  • Source: https://ushe.edu/board-of-higher-education-launches-ai-task-force-begins-statewide-effort-to-expand-ai-workforce-credential/
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: Medium
  • How stable this reading is: Low-medium
  • Last reviewed: 2026-05-13
  • Next planned review: 2026-08-15
  • Themes: education, labor, state-capacity, dispersion, civic-infrastructure
  • What happened: Utah launched a statewide AI task force and a university-issued AI Workforce Credential. More than 50,000 graduates from Utah’s public colleges and universities in the classes of 2025–2027 are expected to be eligible to earn it at no cost beginning July 1, 2026.
  • Why this belongs in the scorecard: This is an institutionally significant statewide public higher-education distribution channel for AI literacy and workforce adaptation.
  • Why this points toward dispersion: Public colleges and universities become a broad AI-capability distribution mechanism rather than leaving AI training primarily to elite firms, private bootcamps, or self-directed learners.
  • Concentration risks / limits: The credential could become narrow employer-aligned compliance training, credential inflation, or a weak symbolic credential if not connected to real skills and opportunities.
  • What could change this reading: Strong participation, portable skills, employer recognition, and broad accessibility would strengthen dispersion. Narrow employer capture, low uptake, or weak labor-market value would weaken it.

5. OpenAI launches the OpenAI Deployment Company

  • Source and date: OpenAI, May 11, 2026; Reuters, May 11, 2026
  • Where this happened: Company / enterprise AI deployment
  • Sources:
  • Initial reading: Concentration +3
  • Scorecard reading: Concentration +3
  • Confidence in this reading: High
  • How stable this reading is: Medium-high
  • Last reviewed: 2026-05-13
  • Next planned review: 2026-08-15
  • Themes: concentration, market-structure, labor, data, public-sector
  • What happened: OpenAI launched the OpenAI Deployment Company, backed by more than $4 billion in initial investment, and agreed to acquire Tomoro, adding about 150 forward-deployed engineers and deployment specialists to help organizations redesign workflows around frontier AI.
  • Why this belongs in the scorecard: This is a major institutional development in the AI deployment layer. It is not merely model access; it is workflow transformation, consulting, engineering, and adoption capacity organized around a frontier model company.
  • Why this points toward concentration: Enterprise and possibly public-sector AI adoption capacity becomes more tightly bundled with a frontier model provider and its partner ecosystem. This may deepen dependence on a small number of AI firms not only for models, but for organizational redesign.
  • Why the story is not one-sided: The deployment company could standardize practices that later diffuse through smaller firms, open tools, or public-sector learning. It may also help organizations adopt AI more effectively than they otherwise could.
  • What could change this reading: Evidence of open standards, portability, public-interest deployments, small-organization access, or broad diffusion of transferable practices could weaken the concentration score. Exclusive partnerships, deep workflow lock-in, proprietary organizational redesign, or expansion into public-sector operating systems would strengthen it.

6. Florida creates state guardrails for large data centers

  • Source and date: Government Technology / Tampa Bay Times, May 8, 2026
  • Where this happened: Florida / state regulation / compute infrastructure
  • Source: https://www.govtech.com/policy/florida-bill-creates-data-center-guardrails-for-state
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: High
  • How stable this reading is: Medium-high
  • Last reviewed: 2026-05-15
  • Next planned review: 2026-08-15
  • Themes: compute, regulation, state-capacity, public-infrastructure, democratic-accountability
  • What happened: Florida enacted Senate Bill 484 to regulate large-scale data centers. The law requires water-management districts to deny permits where proposed water use would harm local water resources, requires reclaimed water where possible, and requires large data centers to bear their own full cost of electric service rather than shifting those costs to the public.
  • Why this belongs in the scorecard: The governor signed the bill into law. This is no longer only a proposal, study, or public controversy.
  • Why this points toward dispersion: State government is asserting public-interest conditions over AI-related physical infrastructure. The law does not stop data-center development, but it makes water use, resource impacts, and ratepayer exposure part of the governance structure rather than treating them as private-site issues.
  • Concentration risks / limits: The law could be implemented narrowly, and an earlier provision restricting nondisclosure agreements between government officials and developers was reportedly stripped from the bill.
  • What could change this reading: Active enforcement, replication by other states, or use of the law to deny or condition major projects would strengthen the dispersion score. Weak implementation or transparency loopholes would weaken it.

7. Reno pauses new data-center applications

  • Source and date: Government Technology / Las Vegas Review-Journal, May 15, 2026
  • Where this happened: Reno, Nevada / local government / compute infrastructure
  • Source: https://www.govtech.com/artificial-intelligence/reno-nev-is-first-in-state-to-pause-data-center-development
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: High
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-15
  • Next planned review: 2026-08-15
  • Themes: compute, local-government, regulation, civic-infrastructure, democratic-accountability, public-infrastructure
  • What happened: Reno became the first local government in Nevada to pause new data-center applications. The city council voted 6-1 to adopt a temporary moratorium on conditional-use permits for data centers, pending further action and possible long-term code amendments. The decision followed a packed special meeting with extensive public comment.
  • Why this belongs in the scorecard: The council vote created an immediate permitting pause. That is a concrete institutional action, not merely opposition, study, or general concern.
  • Why this points toward dispersion: A city is slowing the default buildout path long enough to develop rules, rather than allowing the permitting pipeline to define the policy. The signal is especially important because Northern Nevada is becoming a significant AI/data-center geography.
  • Concentration risks / limits: Existing pipeline projects may absorb most practical impact, and the pause may expire without durable code changes.
  • What could change this reading: Durable code amendments around water, power, ratepayer impacts, public notice, tribal input, or environmental review would strengthen dispersion. Expiration without meaningful rules would weaken it.

8. Utah governor adds phased approvals and environmental review for Stratos data center

  • Source and date: Business Insider, May 10, 2026
  • Where this happened: Utah / state-local compute infrastructure
  • Source: https://www.businessinsider.com/utah-data-center-box-elder-kevin-oleary-governor-spencer-cox-2026-5
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: Medium-high
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-15
  • Next planned review: 2026-08-15
  • Themes: compute, regulation, state-capacity, public-infrastructure, local-government, democratic-accountability
  • What happened: After local opposition to the Stratos Project in Box Elder County, Utah Gov. Spencer Cox announced new demands for the large data-center project, including new approvals for each phase, an initial cap not to exceed 1.5 gigawatts, review of air-quality permits, and direction to protect water and use environmentally sensitive cooling systems.
  • Why this belongs in the scorecard: The governor announced specific state actions after county approval and public protest. This is more concrete than a general statement of concern.
  • Why this points toward dispersion: State leadership is responding to local concern by inserting staged approval and environmental scrutiny into a massive AI-infrastructure project. It does not reject the project; it converts “approve the megaproject” into a more conditional governance sequence.
  • Concentration risks / limits: The requirements may become mostly procedural if later phases are effectively automatic or if state action preempts stronger local governance.
  • What could change this reading: Enforceable conditions, public disclosure, or meaningful phase-by-phase review would strengthen dispersion. Procedural review without leverage would weaken it.

9. Maryland challenges regional transmission cost-shifting for data centers

  • Source and date: Maryland Office of People's Counsel, May 7, 2026
  • Where this happened: Maryland / regional grid governance / compute infrastructure
  • Source: https://content.govdelivery.com/accounts/MDOPC/bulletins/415c9b6
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: High
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-15
  • Next planned review: 2026-08-15
  • Themes: compute, public-infrastructure, regulation, state-capacity, democratic-accountability
  • What happened: The Maryland Office of People's Counsel filed a complaint at FERC challenging PJM's rules for assigning regional transmission costs driven by data centers. OPC argues that PJM's cost-allocation method unfairly assigns Maryland customers responsibility for $2 billion in capital expenditures and would increase Maryland customer bills by about $1.6 billion over ten years.
  • Why this belongs in the scorecard: A formal complaint has been filed at FERC. It is a concrete regulatory action with a docketed institutional pathway.
  • Why this points toward dispersion: A state consumer advocate is contesting hidden infrastructure subsidies for AI compute buildout. The issue is not whether data centers exist, but who pays for the grid expansion they require and whether ordinary ratepayers subsidize concentrated compute capacity located elsewhere.
  • Concentration risks / limits: FERC may reject the complaint, or the dispute may remain isolated rather than prompting broader governance reform.
  • What could change this reading: FERC cost-allocation changes, similar complaints by other states, or redesign of large-load tariffs around cost causation would strengthen dispersion. Rejection without broader uptake would weaken it.

10. California uses AI-supported public engagement on AI's labor impacts

  • Source and date: Government Technology, May 12, 2026
  • Where this happened: California / public engagement / workforce policy
  • Source: https://www.govtech.com/artificial-intelligence/in-the-age-of-ai-a-new-approach-to-public-engagement
  • Initial reading: Dispersion +1
  • Scorecard reading: Dispersion +1
  • Confidence in this reading: Medium-high
  • How stable this reading is: Low-medium
  • Last reviewed: 2026-05-15
  • Next planned review: 2026-09-30
  • Themes: civic-infrastructure, labor, democratic-accountability, public-sector, state-capacity, data
  • What happened: California's Office of Data and Innovation is using Engaged California to gather public input about how AI is affecting work and the economy. The process uses AI to synthesize and categorize comments, includes a later live-forum phase, and is expected to produce a public report in September. ODI also emphasizes open-sourcing project data, code, and methodology.
  • Why this belongs in the scorecard: The engagement process is live and institutionally sponsored by a state digital-service office. It is more than a concept note or conference discussion.
  • Why this points toward dispersion: This points toward AI-assisted civic infrastructure: using AI not only to automate public administration, but to help public institutions hear, structure, and analyze broad public input.
  • Concentration risks / limits: The process could become shallow, elite-skewed, or a legitimacy exercise if outreach is weak, synthesis is opaque, or policy impact is minimal.
  • What could change this reading: Broad participation, reusable data and methodology, and policy action after the September report would strengthen dispersion. Shallow participation or lack of influence would weaken it.

11. Hill County, Texas adopts one-year rural data-center moratorium

  • Source and date: Texas Tribune, May 12, 2026; AP, May 14, 2026; Tom’s Hardware, May 16, 2026
  • Where this happened: Hill County, Texas / county government / compute infrastructure / rural land use
  • Sources:
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: Medium-high
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-18
  • Next planned review: 2026-08-15
  • Themes: compute, local-government, regulation, democratic-accountability, civic-infrastructure, public-infrastructure
  • What happened: Hill County commissioners voted 3-2 to impose a one-year moratorium on new data-center construction in unincorporated areas of the county. Reporting frames the action as the first known temporary data-center moratorium by a Texas county, though state officials and some observers have questioned whether counties have legal authority to impose such bans.
  • Why this belongs in the scorecard: The county commission has taken a concrete governance action. The moratorium may be legally challenged, but the vote itself changes the local approval path and creates a one-year pause unless overturned or lifted.
  • Why this points toward dispersion: A rural county is trying to slow AI/data-center buildout long enough to study public-health, public-safety, infrastructure, utility, and land-use impacts before permitting becomes irreversible. The signal is especially important because developers may seek rural and unincorporated locations where city-level oversight is weaker.
  • Concentration risks / limits: Legal uncertainty could make the moratorium short-lived. State preemption, attorney-general intervention, or litigation could convert the episode into a concentration signal by limiting local authority over AI infrastructure.
  • What could change this reading: The dispersion score would strengthen if the moratorium survives legal challenge, leads to enforceable county standards, or inspires similar rural controls. It would weaken if state officials, courts, or the legislature preempt the county’s action.

12. Harlingen, Texas adopts 120-day data-center moratorium

  • Source and date: Government Technology / Valley Morning Star, May 22, 2026
  • Where this happened: Harlingen, Texas / local government / compute infrastructure / water, energy, land use
  • Source: https://www.govtech.com/artificial-intelligence/south-texas-city-sets-moratorium-amid-data-center-interest
  • Initial reading: Dispersion +1
  • Scorecard reading: Dispersion +1
  • Confidence in this reading: High
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-22
  • Next planned review: 2026-08-22
  • Themes: compute, local-government, regulation, public-infrastructure, civic-infrastructure, democratic-accountability
  • What happened: Harlingen city commissioners adopted a 120-day moratorium on applications for data-center projects while the city studies potential impacts on water, energy, wastewater, electrical coordination, transportation, emergency services, land-use compatibility, noise, mechanical/electrical infrastructure, and the city's comprehensive plan. The move follows regional data-center interest in Cameron County, including pre-permitting discussions with two companies and a prior proposal for a 1,785-acre project near Valley International Airport.
  • Why this belongs in the scorecard: This is a formal local-government action, not merely public concern, coalition pressure, or early discussion. The moratorium creates an immediate temporary permitting pause.
  • Why this points toward dispersion: Harlingen is acting before a formal application becomes a fait accompli. The city is using a short moratorium to create time for zoning, infrastructure, environmental, and utility review before high-load compute development enters the local approval path. The action also strengthens the broader pattern of Texas localities and counties seeking more public leverage over data-center siting, water, energy, and local-resource burdens.
  • Concentration risks / limits: The moratorium is short, no formal city application appears to be pending, and the most active development interest may sit in county jurisdiction rather than inside Harlingen. Projects could proceed outside city limits if county authority remains weak, or state preemption could limit local control.
  • What could change this reading: The score would strengthen if Harlingen adopts enforceable data-center zoning, water, energy, infrastructure, disclosure, or emergency-services standards. It would weaken if the moratorium expires without policy changes, if projects simply proceed outside city limits without county authority, or if state preemption limits local control.

13. California signs executive order on AI workforce disruption and benefit-sharing

  • Source and date: Governor of California, May 21, 2026; StateScoop, May 21, 2026; CalMatters, May 22, 2026
  • Where this happened: California / state government / workforce / small business / civic infrastructure
  • Source: https://www.gov.ca.gov/2026/05/21/governor-newsom-signs-first-of-its-kind-executive-order-to-prepare-workers-and-businesses-for-potential-ai-disruption/
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: Medium-high
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-26
  • Next planned review: 2026-08-31
  • Themes: state-capacity, labor, workforce, transition-shock, civic-infrastructure, implementation-capacity, public-sector, dispersion
  • What happened: California Governor Gavin Newsom signed an executive order directing state agencies to confront AI's economic effects on workers and small businesses. The order creates an agency work program around labor-market shifts, worker early-warning systems, small-business adaptation, and possible benefit-sharing mechanisms so AI-generated wealth does not accrue only to major technology companies.
  • Why this belongs in the scorecard: This is a signed executive order, not a proposal or coalition request. It creates a concrete state-agency work program, deadlines, and a policy-development channel around AI disruption.
  • Why this points toward dispersion: California is beginning to build public transition machinery around AI rather than leaving labor-market adjustment entirely to private firms, workers, and markets. The order is a state-capacity and civic-infrastructure signal because it treats AI disruption as a governable public problem involving workers, small businesses, early-warning systems, and potential benefit-sharing.
  • Concentration risks / limits: The order could remain largely procedural if it produces only studies, voluntary guidance, or weak recommendations. It could also be captured by incumbent firms if policy design focuses mainly on helping businesses adopt AI without meaningful worker protections or public accountability.
  • What could change this reading: The score would strengthen if California produces concrete WARN Act updates, worker-transition funding, small-business implementation support, benefit-sharing mechanisms, or durable public reporting. It would weaken if the order produces only reports with little legislative or agency follow-through.

14. Groupon launches explicitly AI-native restructuring and cuts up to 400 jobs

  • Source and date: Wall Street Journal, May 26, 2026; company 8-K summaries, May 26, 2026
  • Where this happened: Company / local commerce marketplace / workforce restructuring / business model redesign
  • Source: https://www.wsj.com/business/groupon-to-cut-nearly-a-quarter-of-workforce-in-restructuring-a286bcd5
  • Initial reading: Concentration +1
  • Scorecard reading: Concentration +1
  • Confidence in this reading: Medium
  • How stable this reading is: Low-medium
  • Last reviewed: 2026-05-26
  • Next planned review: 2026-08-31
  • Themes: labor, workforce, workforce-restructuring, lean-firm, market-structure, workflow-redesign, business-adoption, vendor-dependence, concentration
  • What happened: Groupon announced plans to lay off up to 400 employees, nearly one-quarter of its global workforce, as part of a restructuring aimed at transforming the company into an AI-native business. Reporting indicates the company expects up to $25 million in annualized savings, plans to reinvest part of the savings into marketing, AI infrastructure, and talent, and raised its full-year adjusted EBITDA forecast.
  • Why this belongs in the scorecard: The restructuring is concrete, public, and company-approved. Unlike an ordinary layoff story, the stated AI-native transformation gives the item a clear AI mechanism, though the score should remain modest until operating results are visible.
  • Why this points toward concentration: Groupon is not a frontier AI platform, but the case suggests that AI-linked labor-light restructuring is spreading beyond mega-platforms into older digital intermediaries and marketplace businesses. It supports the lean-firm hypothesis as a pattern to test: firms may try to maintain or expand output with fewer employees by rebuilding workflows around AI.
  • Why the story is not one-sided: If Groupon's AI-native restructuring helps a weakened incumbent become more competitive against larger platforms, the case could become mixed. It might also reveal practices that later diffuse to smaller firms.
  • What could change this reading: The concentration score would strengthen if Groupon shows sustained revenue, margin, or output gains with lower headcount, discloses AI systems replacing specific roles, or becomes an example other marketplace firms copy. It would weaken if the cuts prove to be ordinary turnaround cost-cutting with AI used mainly as branding.

15. Jefferson County, Colorado adopts 10-month data-center moratorium

  • Source and date: Jefferson County, May 19, 2026
  • Where this happened: Jefferson County, Colorado / county government / compute infrastructure / land use
  • Source: https://www.jeffco.us/CivicAlerts.asp?AID=2596
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: High
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-29
  • Next planned review: 2026-08-31
  • Themes: compute, local-government, regulation, civic-infrastructure, democratic-accountability, public-infrastructure, dispersion
  • What happened: The Jefferson County Board of Commissioners voted during a May 19 public hearing to impose a 10-month moratorium on data centers. The moratorium applies to new applications for data-center development and rezoning to allow data centers, while excluding land already specifically zoned to allow data centers through a Planned Development if the facility is at least 1,500 feet from any dwelling.
  • Why this belongs in the scorecard: This is a formal county action, not merely public opposition, a staff study, or a proposal. It creates an immediate local permitting pause while the county considers how data centers should be governed.
  • Why this points toward dispersion: This is a strong Colorado local-governance signal. After Colorado's statewide data-center bills failed, Jefferson County is joining Denver and Larimer County in using local land-use authority to create a governance pause. The dispersion value is procedural: county officials are slowing the default buildout pathway before private permitting momentum becomes irreversible.
  • Concentration risks / limits: The moratorium includes an exemption for land already specifically zoned for data centers under certain conditions. The pause could also expire without durable rules, or development could shift to nearby jurisdictions with weaker review.
  • What could change this reading: The score would strengthen if Jefferson County adopts enforceable standards around power, water, noise, air quality, fire risk, setbacks, public disclosure, ratepayer exposure, or community-benefit requirements. It would weaken if the moratorium expires without durable rules, if major projects are exempted, or if development simply shifts to nearby jurisdictions with weaker oversight.

16. Ohio pauses new data-center tax exemptions during legislative review

  • Source and date: Ohio Governor Mike DeWine, May 27, 2026; Associated Press, May 28, 2026
  • Where this happened: Ohio / state economic-development policy / compute infrastructure / tax incentives
  • Source: https://governor.ohio.gov/wps/portal/gov/governor/media/news-and-media/governor-dewine-announces-pause-of-data-center-tax-exemption
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: High
  • How stable this reading is: Medium
  • Last reviewed: 2026-05-29
  • Next planned review: 2026-08-31
  • Themes: compute, regulation, state-capacity, public-infrastructure, market-structure, democratic-accountability, dispersion
  • What happened: Ohio Governor Mike DeWine directed the Ohio Tax Credit Authority to pause consideration of new data-center tax exemptions while the legislature studies data-center growth. The governor's office emphasized that the move is not a data-center ban; it suspends the ability of data centers to request new exemptions while policymakers review the incentive structure.
  • Why this belongs in the scorecard: This is a governor-directed administrative pause tied to a legislative study process. It is a concrete state-capacity action involving the public-subsidy layer of AI compute infrastructure.
  • Why this points toward dispersion: This moves the governance question from siting alone to public subsidy design. The key issue is whether AI compute firms receive public tax advantages while imposing power, water, grid, and fiscal burdens. Pausing new exemptions creates leverage to reassess whether public incentives are aligned with public benefits.
  • Concentration risks / limits: The pause could be short-lived and followed by resumed subsidies with little change. A weak review could legitimize the existing incentive structure rather than reforming it.
  • What could change this reading: The score would strengthen if Ohio reforms tax exemptions, imposes cost-causation rules, adds ratepayer protections, requires disclosure, or conditions incentives on public benefits. It would weaken if the pause ends quickly and incentives resume without meaningful changes.

17. Huron County, Michigan adopts three-year data-center moratorium

  • Source and date: Huron Daily Tribune, May 27, 2026
  • Where this happened: Huron County, Michigan / county government / rural land use / compute infrastructure
  • Source: https://www.michigansthumb.com/news/article/huron-county-data-center-moratorium-22278673.php
  • Initial reading: Dispersion +2
  • Scorecard reading: Dispersion +2
  • Confidence in this reading: Medium-high
  • How stable this reading is: Medium-low
  • Last reviewed: 2026-05-29
  • Next planned review: 2026-08-31
  • Themes: compute, local-government, regulation, civic-infrastructure, public-infrastructure, democratic-accountability, dispersion
  • What happened: Huron County commissioners approved a three-year moratorium on data centers and similar projects. The moratorium halts land-use permits while the county studies and develops standards. Commissioners adopted a longer pause than legal counsel recommended, following community opposition focused on farmland, environmental protection, and local economic impacts.
  • Why this belongs in the scorecard: The county commission has adopted the moratorium. This is a concrete land-use governance action rather than an early-stage proposal.
  • Why this points toward dispersion: This is another local-government catch-up signal, but with a longer and potentially more legally contestable pause. It suggests that some rural governments are no longer treating data centers as ordinary industrial development and are trying to create time for public standards before projects advance.
  • Concentration risks / limits: The three-year duration may invite legal or political challenge. If the moratorium is struck down or shortened, the episode could instead reveal the limits of local authority over AI compute infrastructure.
  • What could change this reading: The score would strengthen if the county develops legally durable standards that survive challenge and influence other rural jurisdictions. It would weaken if the three-year period is struck down, shortened under pressure, or used only as a symbolic delay.

Read more