Do communities and public institutions gain lasting skills and resources?
Direction of power
Leaning dispersion
Window
Opens
Confidence
Medium-high
Movement since August
Shared institutions, literacy and procurement support expand
End-of-September assessment
September expands institutions that can make AI useful and governable outside large private firms. NAIRR's federally funded operating center pools access and support for researchers and educators. Canada's Amii initiative funds broad literacy; Mitacs placements are a separate route to project expertise. Schools' pauses, contractual standards and participation rules also preserve public authority over the purposes and terms of use.
This continues July and August's movement toward reusable public capacity. The critical distinction is between receiving a tool and possessing the capacity to choose, evaluate, procure and govern it. The GSA agreement expands access and support but begins in October and can increase dependence on OpenAI. Corporate and philanthropic programmes are useful access routes whose durability, curriculum independence and recipient control remain questions. The Gates Foundation commitment is retained in the approved record, but its thinner source entry should not carry the lane assessment by itself.
The direction remains leaning dispersion because funded public connective institutions and broad education mechanisms extend capacity beyond isolated subscriptions. September does not establish that every small government or underserved community receives comparable support, nor that public participation has become decisive in AI investment choices.
Other evidence. The standing public-capacity review below examines the seven-part public layer and the two-way role of municipal networks. California's September 9 AskCA pilot invites resident testing and uses public-source evaluation, while depending on Anthropic's model. A planned September 30 jobs feature is not independently verified as delivered. African public-official training remains an unscored selected-applicant pathway with actual course access unresolved. California AB2392 establishes higher-education procurement and training work, not delivered systemwide standards.
What would change this assessment. Operating shared services with broad uptake; evidence of retained public skills and vendor-changing collective requirements; accessible training and public inventories; effective community influence over problem definition; and transparent procurement, evaluation, remedy and exit terms.
Further public-readiness evidence. SUNY's September 24 account describes faculty support, an open educational resource and fall course integration. Zoom's September allocation supports nonprofit intermediaries within an existing commitment, rather than a new pledge or independently evaluated outcomes. Singapore's September 24 financial-sector Co-Lab involves employers and unions and targets a combined workforce exceeding 80,000 by 2028; a target population is not a completed training count. America.gov's September 29 order and information launch establish a federal interface architecture, with actual transaction integration still prospective. These remain unscored supporting cases or existing Watchlist items. See the unscored evidence register for dates, retained dispositions and source links.
Supporting scored events
The complete relevant event list follows. Primary and secondary references preserve one underlying event and score. Source holds and later commencement dates remain explicit; publication does not resolve them.
September scored actionSeptember 30, 2026 evidence cutoff
New York City Public Schools established a one-year 2026–27 moratorium on student-facing generative AI for students in 2-K through eighth grade, affecting nearly 600,000 students. Companion chatbots are prohibited across grades; approved teacher uses remain available; high-school students receive AI critical-thinking instruction and limited controlled pilots. A Technology in Schools Coalition of students, educators, parents, elected officials, advocates, unions and experts will assess the policy and pilots during the school year.
China’s Cyberspace Administration of China reported the second phase of its “Clear and Bright — Rectifying Chaos in AI Applications” campaign, including removal of more than 5.61 million pieces of unlawful or non-compliant AI-related content, action against more than 49,000 accounts and 2,400 websites/apps, and coordinated tightening of controls by major platforms, model providers, and app stores.
LAUSD blocked generative-AI tools for all K–12 students on district devices while it reviews instructional uses and safeguards, reversing a prior policy that allowed students age 13 and older to use generative AI after digital-citizenship training.
Action/date record: September 7, 2026; publicly announced September 9, 2026. Approved score: Dispersion +1. Primary lane: 1. Secondary lanes: 6.
The American Federation of Teachers’ National Academy for AI Instruction and Microsoft signed a binding agreement establishing reusable AI safety and privacy terms that Microsoft must make available to education customers for incorporation into district agreements. The provisions address training on student/educator data, tracking and repurposing, transparency, human oversight, audit and remedy. Districts must affirmatively incorporate the terms before they become enforceable in a district’s own contract, and the standard applies to defined educational AI products rather than every general-purpose Microsoft product used by schools.
California enacted a coordinated child-technology package. SB 1119 (“Adam’s Law”) requires companion-chatbot operators serving children to assess child-safety risks before releasing a new or substantially modified system, mitigate identified risks, implement crisis-response protocols and parental controls, restrict specified manipulation and advertising, provide incident reporting, and undergo independent child-safety audits subject to Attorney General access and enforcement. AB 1159 strengthens protections for student information used in AI systems, including restrictions on using covered student data to train generative AI or develop AI systems. SB 867 temporarily prohibits manufacturing or selling companion-chatbot toys designed for children under 16 through January 1, 2031.
Taiwan’s Administration for Cyber Security, Ministry of Digital Affairs, formally established a cross-government Task Force on Frontier AI Cybersecurity Risks and adopted a phased policy covering frontier-AI cyber-risk monitoring, shared defensive capability, secure-by-design and supply-chain procurement requirements, threat-intelligence coordination, and longer-term domestic AI cybersecurity capacity across government, industry, academia, and research institutions.
Tucson’s new large-scale data-center regulations became effective after a 6–1 Mayor and Council adoption on August 5. Large projects now face neighborhood coordination, a public hearing, Zoning Examiner review and final Mayor and Council consideration, together with requirements addressing water and power adequacy, energy-mix disclosure, environmental review, noise, enhanced setbacks and potential off-site sensitive-land protection.
Qualcomm and ASUS launched a Pharmaceutical AI Agent field-deployment program for more than 50 demonstration pharmacies across southern Taiwan. The system uses the open-weight gpt-oss-20b model, public Taiwan FDA drug information, OCR and retrieval components, and local/on-premises inference on donated AI endpoint devices to support medication-safety and pharmacy workflows.
Limits and interpretation: Do not describe the deployment as compressing gpt-oss-120b into 20B. OpenAI released gpt-oss-120b and gpt-oss-20b as distinct models; the Taiwan program specializes and deploys the 20B variant. Field validation is beginning, demonstrated performance at scale remains limited, and the deployment still depends materially on Qualcomm hardware and ASUS/Taiwan implementation infrastructure.
NSF awarded $35 million over five years to the San Diego Supercomputer Center, working with the Texas Advanced Computing Center, to establish the National Artificial Intelligence Research Resource Operations Center. The center will operate the national portal, integrate computing, data, models and tools from public and private providers, and provide support, training and coordination as NAIRR transitions from a pilot toward sustained infrastructure.
Limits and interpretation: A substantial portion of frontier capability beneath that public interface remains supplied by a concentrated commercial industry. NAIRR therefore increases contestability without eliminating dependence on concentrated upstream providers or proving full workload portability among them.
OpenAI committed $1 billion in subsidized Daybreak cybersecurity access, training, technical assistance and partnerships for state and local governments, utilities, health organizations, community banks, nonprofits and other essential-service defenders, including a U.S. pilot with MS-ISAC.
The Government of Canada launched a $13 million National AI Literacy Initiative with the Alberta Machine Intelligence Institute (Amii), providing free AI education for students, educators, workers and communities. The program targets up to one million post-secondary students and more than 50,000 K–12 educators and builds on an existing Amii consortium of 61 post-secondary institutions representing more than 900,000 students. Consortium expansion opened September 10, with course delivery and additional national modules scheduled to begin during September and the fall.
GSA and OpenAI executed a 27-month OneGov agreement, effective October 1, 2026 through December 31, 2028, that makes ChatGPT available to eligible federal, state, local and tribal governments with a 50% discount on consumption-based model use, no platform-access fee, no minimum order or spending commitment, multiple purchasing routes, government-data safeguards, training and deployment support. The agreement also includes most-favored-pricing protections, utilization and savings reporting, and a one-year checkpoint.
Limits and interpretation: The agreement was executed in September but takes effect October 1. September status is adopted procurement access, not completed implementation.
Canada’s federal government announced up to C$162 million over five years, beginning in 2026–27, for Mitacs AI Advantage. Evan Solomon, Minister of Artificial Intelligence and Digital Innovation, announced the funding in a release issued by Innovation, Science and Economic Development Canada. The money supports 10,000 co-funded work-placement opportunities for post-secondary students, recent graduates and postdoctoral fellows. Participants will work with Canadian businesses to help them adopt AI and develop innovations, while gaining paid practical experience themselves. Mitacs delivers the program and connects academic talent with organizations that need project expertise.
The practical mechanism is a subsidized placement working on an organization’s project, rather than simply access to an AI course or software subscription. The operational program accepts applications for adoption and research projects. Matching contributions, eligibility requirements and application review apply; the terms inspected during the September 26 review also include eligible municipal, nonprofit and Indigenous-government partners. The central federal announcement emphasizes Canadian businesses. The 10,000 placements are a program target, not a count of completed placements, and the funding ceiling is not evidence that all money has been disbursed.
GSA’s USAi platform completed its transition from a free pilot into a paid cost-recovery shared service for federal agencies. USAi is already used by 25 agencies, with 16 more expected to onboard by year-end, and provides a shared Chat interface, unified API and administrative/evaluation Console across models from multiple providers. Agency charges combine a platform fee with model usage passed through at cost.
Limits and interpretation: The service remains federal-only, agencies must now fund their own participation, and substantial authority over provider selection and platform architecture is concentrated in GSA. The new cost-recovery model could also produce uneven adoption across agencies with different budgets and implementation capacity.
Gov. Maura Healey signed Executive Order 658, making state permitting for data centers above 25 MW contingent on conformity with Massachusetts's responsible-development framework and a community-benefits agreement. The order directs MassDEP to establish protocols requiring enough incremental clean generation to cover annual project consumption and, by December 31, an alternative-compliance-payment mechanism whose proceeds flow to a Ratepayer Protection Fund to reduce electricity-supply costs for other customers. It also directs state support for municipal project review, large-load cost protection, queue discipline and greater transparency.
Limits and interpretation: Some downstream mechanisms still require rulemaking and DPU action, but the permit gate and community-benefits requirement are operative and MassDEP has a mandatory December 31 implementation deadline for the alternative-compliance mechanism.
Action/date record: September 2026; processed September 15. Approved score: Dispersion +2. Primary lane: 6. Secondary lanes: 7.
The foundation committed $1 billion to help lower-resource institutions and communities build and use AI capacity. The size, explicit equity purpose and implementation orientation make this more than a general pledge, while provider dependence and the durability of recipient capability remain follow-up questions.
Source hold: The approved ledger refers to an alert source archive, but an operative URL was not recovered from the records materialized for this review. The score is preserved; actor, precise action date, text and scope need source-archive reconciliation before this becomes fully verified public evidence.
September scored actionSeptember 30, 2026 evidence cutoff
Spain launched IA360 as a twelve-month government program combining AI oversight, public-interest deployment measures, business support, child protections and a formal process involving employers, unions and government in a proposed national AI social compact.
Verizon launched a national free-training initiative combining $50 million new funding with its existing $20 million Reskilling and Career Transition Fund. Multi-provider training is paired with community nonprofit support for workers, job seekers, educators and small businesses.
Limits and interpretation: The $70 million headline is not all new spending. Corporate sponsorship, curriculum independence, durable delivery, participation and transferable outcomes require follow-through. Training access is not itself worker decision authority.
Action/date record: September 30, 2026. Approved score: Dispersion +1. Primary lane: 1. Secondary lanes: 6. Stable record identity: D1001-25.
SB1159 was signed September 30. Public-records, open-meeting, administrative, CEQA and coastal participation rules do not treat autonomous AI or robots as independent persons, and the law prohibits falsely claiming a natural person participated. It expressly preserves human AI assistance at ordinary volume/frequency and disability-related assistive use. It is not a blanket ban on AI-written comments.
Limits and interpretation: An adopted human-participation boundary preserves civic decision space against fabricated participation while retaining legitimate assistance. The +1 recognizes specified public-process protections; agency filtering could itself restrict access and must be assessed.
Queensland commits A$750,000 to ARM Hub’s manufacturing AI uplift programme, with expressions of interest open until October 30. Flagship government contributions are A$150,000–250,000 each; 80% government/20% participant cash plus staff contribution, with at least one regional flagship. Support is delivered as engineering/services rather than cash grants, including data foundations, private Australian-cloud workspaces, integration and support. Eligibility includes 5–500 FTE, at least 50% Queensland-based employees and two years’ Queensland presence. The AI roadmap pathway does not require matched cash and is supported through the funded AI Adopt Centre.
Limits and interpretation: Committed finance and an available application pathway establish a concrete capability-building mechanism for smaller manufacturers. The +1 does not await selections, service commencement or measured skill retention. It is distinct from existing manufacturing microcredentials and is not a government public-service deployment.
Action/date record: September 2, 2026 reporting; formal plan entered into force August 18, 2026. Approved score: Dispersion +1. Primary lane: 7. Secondary lanes: 6, 1.
Türkiye’s 2026–2030 Artificial Intelligence Action Plan entered into force through Presidential Circular 2026/9 and establishes a government-wide implementation architecture including mandatory AI training with assessed micro-certificates, senior institutional AI administrators, three-year agency roadmaps, an Identify–Pilot–Scale cycle, at least 30 pilots across 10 ministries, pilot-to-scale pathways for critical public services, and certification/evaluation for high-impact systems.