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

Policy tool · See record

Worker consultation for major AI deployment

What the tool does

Give workers or their representatives a defined opportunity to influence major AI deployment before choices become fixed. Consultation should identify the decision at issue, provide relevant information and explain how worker input changes the resulting plan.

Current status

A specific participation model has been adopted, but a broad general U.S. worker-consultation rule has not been verified here. California’s higher-education working group includes labor representatives; its convening, recommendations and implementation remain prospective. The worker-protection package is related, but does not itself establish a general consultation entitlement. A separate British Columbia film contract establishes qualified consultation and collective-bargaining mechanisms within covered productions; it is not a general U.S. rule.

Where it has been used

California’s higher-education working group

Added by

Signed AB2392 requires CCC and CSU participation and requests UC participation in a cross-system working group with students, labor, privacy and civil-society representatives. Procurement and training recommendations are due January 1, 2028. Until standards are adopted, systemwide AI contracts require reports within 60 days. A mandated work programme and reporting duty are distinct from delivered training or established procurement standards.

Read more about this case →

California’s worker AI protections

Added by

California’s September 30, 2026 worker package establishes human corroboration and information safeguards for covered automated employment decisions, AI-related restructuring reports and specified surveillance limits. Coverage, public enforcement and exceptions qualify the protections. The package does not create a general worker veto over every AI deployment.

Read more about this case →

British Columbia’s film-industry agreement

Added on

The British Columbia film agreement, ratified February 26, 2025 and effective April 1, 2025 to March 31, 2028, provides qualified employee consultation when AI changes production time. Article 17 keeps covered AI work within the agreement, limits required prompts that displace covered employees, and establishes meetings and training arrangements. Arbitration cannot enjoin AI use under the article. Sideletter 21 preserves legally required effects bargaining with a temporary testing exception.

Read more about this case →

The Inquirer’s negotiated AI protections

Added on

Inquirer workers ratified an agreement on October 2, 2026 with reported AI participation provisions, Guild representation, consent requirements and grievance or severance remedies. The complete operative agreement’s coverage and precise meeting terms are not established in the available account. The report does not establish a universal employee veto or a guarantee against layoffs.

Read more about this case →

Melbourne’s AI disclosure bargaining

Added on

The National Tertiary Education Union’s October 8 bargaining update explains a more precise issue at the University of Melbourne. The union seeks disclosure of AI use in human-resources decisions affecting staff, including recruitment, major workplace change, promotion, discipline and performance.

The university had objected that broader wording could require disclosure for almost any decision, including routine research used when drafting an email. The union clarified the intended employment-decision scope, and the university agreed to work on wording addressing that intent without creating unreasonable administrative burdens. The update does not establish agreement on a binding AI clause. Union bargaining account.

Read more about this case →

British technology workers’ bargaining proposals

Added on

UTAW's October 2 inquiry proposes consultation and collective negotiation before deployment, human review of employment decisions and training that preserves junior learning. Its self-selected interviews describe experiences without measuring population effects. The September TUC motion directs campaigning and model-agreement drafting; employer duties have not been established. Inquiry; TUC motion.

Read more about this case →

Norway’s joint AI implementation forum

Added on

Norway’s government and eight national employer/worker organizations signed an AI cooperation agreement on October 2, recovered through October 15 reporting. It runs through December 31, 2029 and addresses productivity, worker skills, work-environment risks and sharing implementation knowledge. It places adoption within Norway’s existing labour law and codetermination arrangements.

The government’s account establishes a coordination group with the parties, relevant agencies and research bodies; the digitalisation ministry supplies the secretariat and leads follow-up. This is an adopted cooperation arrangement, but the evidence does not establish a new general worker veto, a funded service for every workplace or new enforceable AI clauses in individual contracts.

Read more about this case →

Yale’s technological-change agreement

Added on

Yale workers rallied October 15 during contract bargaining, alongside release of Local 34’s Artificial University report. The union describes pressure to use AI, errors that add work, and jobs disappearing through attrition rather than a single layoff announcement. Its March 2025 membership survey received 2,455 responses; 719 respondents reported AI or automation in their departments. That is not a 2026 survey of all university workers. Union report page; October 16 reporting.

The existing 2022 agreement, page 38, is more precise than a universal mandatory 90-day notice rule. For technological changes with defined job or work consequences, Yale undertakes to endeavor to provide 90 days’ notice. The union can request discussion and effects bargaining. Yale retains implementation authority if agreement fails; the union can arbitrate effects and remedies rather than veto the decision. Whether Yale breached these terms in deploying AI remains an allegation, not an adjudicated finding.

Two qualifications matter. The report acknowledges a hiring freeze and retirement incentives, so falling membership does not isolate AI’s causal contribution. And Yale’s August 28, 2024 announcement committed more than $150 million over five years across computing, tools, hiring and research—not $150 million spent solely on the Clarity chatbot.

Read more about this case →

Singapore’s union–business AI programme

Added on

Singapore’s National Trades Union Congress and Association of Small and Medium Enterprises launched the SME Transformation Accelerator on October 14. ASME would group firms with similar operational problems, help them adopt technology and redesign jobs, and arrange reskilling. The programme targets 120 firms across retail, professional services and logistics, with projected benefits for at least 480 workers. The first planned cohort is 40 retail firms by November; those targets are not delivered results.

Funding uses the existing Company Training Committee grant, not a newly announced programme-sized pot of money. Grant terms allow support for up to 70% of qualifying project costs, require a Company Training Committee and endorsed transformation plan, and attach worker-outcome commitments. Government bodies and their wholly owned subsidiaries are excluded. The launch account is from the participating organizations; independently measured wage or implementation gains are not yet available.

Pooling projects could lower the practical minimum scale for adoption. The worker committee and grant conditions provide an institutional channel beyond simply handing firms software, although they do not establish a general worker veto over AI.

Read more about this case →

Current Affairs negotiates AI production limits

Added on

Current Affairs staff announced a ratified and signed two-year union contract on October 2. The union and employer account describes limits on generative AI writing articles and creating magazine art, worker copyright and a board representative. Helpful uses of technology remain possible. The complete agreement, exact coverage, exceptions and enforcement terms remain uninspected.

Read more about this case →

Oakland technical workers make AI protections part of a threatened strike

Added on

IFPTE Local21 announced a conditional Oakland strike for November5 if bargaining remains unresolved. Workers seek protections concerning responsibility for AI output and workload. These are demands, not a verified signed consultation clause or completed strike.

Read more about this case →

Animation performers ratify contracts with concrete AI protections

Added on

SAG-AFTRA members ratified four-year television and basic-cable animation agreements on October 29. They cover relevant television and new-media animation, not theatrical animated films. The term ends June 30, 2030. The union’s FAQ says most terms start the Sunday after the producers’ association receives ratification notice; specified provisions have other dates.

The union’s agreement summary closes a loophole involving replicas made from existing recordings and prevents later, unprovided scripts from expanding alterations allowed without further consent. It establishes minimum compensation and residuals for specified independently created replicas. Consent for replica-based foreign-language dubbing applies to productions starting July 1, 2027. Transfer provisions preserve contractual responsibility or require an approved transferee’s written commitment, including arbitration rights. Synthetic performers remain possible under qualified notice, bargaining and additional-value provisions; this is not a blanket AI prohibition. Independent ratification report.

Member ratification completes the collective decision. The general contract runs from July 1, 2026 through June 30, 2030, while individual new protections have their own commencement rules. Further notice and final-document checks do not postpone recognition of the ratified agreement.

Read more about this case →

A newsroom contract protects workers’ professional judgment over AI

Added on

Embarcadero Community News Guild members ratified their first three-year contract on October 8. This is older recovered action: the NewsGuild published details October 21, followed by CWA’s October 29 account. The union represents twenty reporters, editors, designers and photo staff at the California nonprofit newsroom.

The union reports that members participate in setting AI policy and cannot be required to use AI in ways conflicting with professional ethics, legal standards, journalistic goals or AI policies. That is narrower than a ban on AI or a worker veto over every deployment. The accounts come from the union family; the complete agreement and clause-specific remedies remain uninspected.

Read more about this case →

NEA publishes model language for bargaining over workplace AI

Added on

The National Education Association’s October 30 model contract language proposes joint labor-management oversight and agreement before AI affecting bargaining-unit employees is adopted. It also proposes educator autonomy, protection against AI-related staffing reductions, restrictions on AI performance evaluation and permission before synthetic use of a person’s identity.

These are terms unions can seek in negotiations. Publishing the template does not mean an employer has accepted them or that educators everywhere now possess those rights. Its six-month reopening provision also requires mutual consent.

Read more about this case →

Maine journalists ratify a human-review requirement for AI content

Added on

News Guild of Maine members ratified an expanded contract on November 12, 2026, according to the Maine AFL-CIO; the guild published its account November 16 and CWA followed November 19. The unified agreement covers nearly two hundred workers, including more than fifty obtaining a first contract. The union accounts describe a duty for human review and approval of AI-generated content before publication.

These are authoritative labor-side accounts of ratification and a specific contract duty, not independent employer confirmation. The full agreement and grievance/remedy text were not retrieved. The verified account is therefore limited to the ratified human-review requirement: it does not establish an individual worker veto, a no-layoff promise or an outsourcing ban.

Read more about this case →

Source snapshot: 2026-11-20. Individual event and verification dates remain as recorded above. Inclusion alone does not assign a score.