The bill amends Washington collective bargaining law for state employees and institutions of higher education, adds definitions related to artificial intelligence, and creates new statutory provisions. It prohibits institutions of higher education and exclusive bargaining representatives from agreeing to contract terms that would prevent implementation of approved affirmative action plans or that are inconsistent with a specified comparable worth agreement from the 1983–1985 biennium. It also lists management rights that employers and institutions may not bargain away, including programs and functions, organizational structure, budget and workforce size (including layoffs), supervision and direction of employees, emergency actions, retirement plans, and health care benefits, with exceptions referenced elsewhere in the act. The bill grants most Washington management service (WMS) members the right to collectively bargain while excluding certain WMS positions (salary bands 3 and 4 and medical band; HR, budget, risk and litigation managers; employees whose primary duties include employee-related investigations; positions reporting directly to specified senior officials; and other cited exclusions). Bargaining over wages for WMS members is limited to WMS salary band levels rather than individual classifications or positions, and only supervisory or nonsupervisory units of WMS salary band 1 and band 2 members within an agency may be designated for bargaining, subject to public employment relations commission review. The governor or the governor’s designee must negotiate for eligible WMS members, and no collective bargaining agreement under the referenced WMS provisions may take effect prior to July 1, 2025. The bill creates new duties in both chapter 41.56 RCW and chapter 41.80 RCW requiring employers to bargain over decisions to adopt or modify uses of artificial intelligence technology when those decisions affect employees’ wages or performance evaluations, while excluding bargaining when the change is a third‑party update to technology already used by employees. It adds or cross‑references definitions for “artificial intelligence” and “third party,” preserves existing contracts until they expire/are renewed/or reopened, and conditions the act’s effectiveness on specific funding being provided by June 30, 2026. Important text is missing from the provided extracts: the full text of the new sections (including the complete definition in RCW 41.80.005), the exceptions in sections referenced as addressing use of technology, and the remainder of a cut‑off provision in Sec. 4, so some implementation details and effective date language are unclear from these materials.

If enacted, state agencies listed (like corrections, social and health services, transportation, health, natural resources, enterprise services, ecology, etc.), their unions, and affected employees will probably face more formal bargaining whenever the employer decides to adopt or change uses of artificial intelligence that affect wages or performance evaluations. That will likely slow some AI-driven rollouts, create additional negotiation costs and delay risks for projects, and give unions a clear trigger to seek compensation or evaluation changes; routine vendor or third‑party software updates generally would not require bargaining. Existing collective bargaining contracts stay in force until they expire, are renewed, or are reopened, and the whole act depends on specific funding being included in the omnibus appropriations by June 30, 2026, so its effect is uncertain until that funding is approved. Washington management service employees and their unions are affected differently: bargaining over wages will be limited to broad WMS salary band levels rather than specific job classifications, only supervisory or nonsupervisory units of WMS salary bands 1 and 2 may be certified (with exclusions for certain higher bands and specified managerial roles), and no new WMS agreements under the section can take effect before July 1, 2025. Institutions of higher education cannot agree to contract terms that would block approved affirmative action plans or conflict with the cited comparable worth agreement, which preserves those obligations but may reduce some bargaining tradeoffs. Important details about the precise definition of AI used here and certain exceptions referenced in other sections are missing from the provided text, so some operational impacts remain unclear.