The Race — Building Democracy in the Age of AI. People race toward a shared finish line.
The Race
← September 2026 overview

SEPTEMBER 2026 · LANE 08

Organizational Forms and Workforce Restructuring

How does AI change work, organizations and workers’ influence?

Direction of power
Leaning concentration
Window
Narrows overall; worker protections Preserve
Confidence
Medium-high
Movement since August
Meaningful worker safeguards adopted; management-led restructuring still dominates

End-of-September assessment

September changes the worker-side institutional position without yet reversing the lane. California's adopted package requires human corroboration for specified automated disciplinary decisions, notice and worker-data access, anti-retaliation protections, displacement reporting and bounded surveillance restrictions. Its strongest employment protections commence July 1, 2027, and enforcement is public rather than a general private lawsuit right. Spain's social-partner process and funded training pathways provide additional counterexamples, with implementation still open.

Management-led organizational recomposition remains the cumulative pattern: task decomposition, compression, selective hiring, outsourcing, redeployment and wider supervision through agents. Existing employer cases do not establish that every layoff was caused by AI or that measured productivity gains explain each restructuring. Worker control over metrics, task design, staffing assumptions and deployment remains less developed than the employer's capacity to reorganize work.

The monthly movement is a meaningful adopted counterweight, not an economy-wide redistribution of bargaining power. The earlier provisional SB947 score is superseded by the coordinated +2 package, not added to it. No scores are proposed for forecasts, surveys, alleged machine-management practices or voluntary principles.

Other evidence. The workforce case file retains Samsung's union allegation about AI-mediated staffing, attributed employer restructuring and union responses without converting them into verified causation. PwC's September 29 survey covers 49,364 workers across 48 countries; reported use and uneven training access are descriptive, self-reported findings. Littler's employer survey informs expected labor-relations issues, not observed displacement. York's conciliation does not establish an AI agreement. The OECD financing chapter explains that a falling labor share can affect social-insurance receipts even without net job losses; that mechanism does not prove AI caused the shift.

What would change this assessment. Binding worker participation in redesign; enforceable access to decision evidence and remedies; implemented safeguards with meaningful public enforcement; measured wages, job quality and task changes; or credible evidence that small independent organizations gain autonomy rather than becoming dependent subcontractors.

Further worker-side evidence. Singapore's Co-Lab combines education and tripartite participation but does not establish new worker decision rights. Kelly's September 24 proposal links AI-related revenues to transition support without appropriating an operating fund. The IMF task-assistance estimate cannot identify who captures value or prove displacement. These counterexamples and analytical limits strengthen the distinction between access to learning and authority over redesign. 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

Canada funds 10,000 co-funded AI work placements through Mitacs

Action/date record: 2026-09-25 announcement; funding begins 2026–27. Approved score: Dispersion +1. Primary lane: 7. Secondary lanes: 6, 8. Stable record identity: D26-01.

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.

Sources: Source 1 ↗; Source 2 ↗; Source 3 ↗.

Open the existing event record →

September scored actionSeptember 30, 2026 evidence cutoff

Spain launches the twelve-month IA360 responsible-deployment plan

Action/date record: September 21, 2026. Approved score: Dispersion +2. Primary lane: 1. Secondary lanes: 6, 7, 8.

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.

Sources: Source 1 ↗.

Open the existing event record →

September scored actionSeptember 30, 2026 evidence cutoff

D1001-22 — California worker package adopts human control, displacement reporting and surveillance limits

Action/date record: September 30, 2026. Approved score: Dispersion +2. Primary lane: 8. Secondary lanes: 1, 7. Stable record identity: D1001-22.

SB947, SB951 and AB1883 were signed on September 30. SB947 establishes human corroboration for discipline or dismissal principally based on an automated decision system, notice and information about a worker’s own data, and anti-retaliation provisions. Its principal employment protections commence July 1, 2027. Public authorities enforce the law; there is no private right of action, and collective-bargaining and federal-law exceptions qualify coverage. SB951 adds AI-replacement information to the final 60-day CalWARN notice, quarterly EDD summaries and a hiring report beginning January 1, 2028. AB1883 restricts specified neural and emotion surveillance with exceptions; it does not prohibit every form of monitoring. AB1331 bathroom privacy is wider digital-surveillance context, not a separate AI score.

Limits and interpretation: The coordinated package gives workers and public enforcers specified human-control, information and privacy mechanisms. The +2 recognizes the combined institutional consequence. It preserves the existing SB947 lifecycle; the earlier provisional +1 is superseded, not added. This is adoption rather than demonstrated compliance or improved labor-market outcomes.

Sources: Source 1 ↗; Source 2 ↗; Source 3 ↗; Source 4 ↗; Source 5 ↗; Source 6 ↗.

Open the existing event record →

September scored actionSeptember 30, 2026 evidence cutoff

D1001-13 — Queensland funds an open-application manufacturing AI service programme

Action/date record: September 30, 2026. Approved score: Dispersion +1. Primary lane: 7. Secondary lanes: 6, 8, 9. Stable record identity: D1001-13.

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.

Sources: Source 1 ↗; Source 2 ↗; Source 3 ↗.

Open the existing event record →

Earlier actions recorded or announced in September

These support the cumulative assessment. They are not new September actions, and their existing scores and ledger placement are preserved.

Earlier action · September recordSeptember 30, 2026 evidence cutoff

DDD Dehradun — operating training-and-employment AI capability center

Action/date record: Operational August 1, 2026; announced September 11; processed September 13. Approved score: Dispersion +1. Primary lane: 7. Secondary lanes: 8.

About 150 hires reportedly onboarded at DDD's first Indian Global Capability Centre. Foundational training lasts two to four weeks, followed by client-specific instruction. Expansion to 800–1,000 remains a target.

Limits and interpretation: DDD has a genuine U.S. 501(c)(3) entity alongside commercial operations and philanthropic support. Nonprofit ownership of every operating entity and freedom from vendor/government influence are not established. Indian ownership chain, client concentration and center-specific funding remain unresolved. Track transferable skills, wages, advancement, external placement and customer dependence, not simply headcount. New-center outcomes remain unproven.

Sources: Source 1 ↗; Source 2 ↗; Source 3 ↗.

Open the existing event record →

Read the additional unscored supporting cases and research →

These findings inform interpretation without receiving event or trend scores.

← All nine lanes