This bill creates a new regulatory chapter in Title 19 RCW and a new chapter in Title 42 RCW, amends 2024 c 163 s 2 (uncodified), and establishes new duties for persons who develop or deploy “high‑risk artificial intelligence systems” in Washington. It defines key terms (including “artificial intelligence,” “high‑risk artificial intelligence system,” “deployer,” “developer,” “algorithmic discrimination,” and “consequential decision”) and requires deployers beginning July 1, 2027 to use industry‑standard risk management frameworks (for example, NIST or ISO/IEC 42001), maintain a risk management program, conduct annual reviews to detect algorithmic discrimination, complete impact assessments for systems and for substantial modifications, retain impact assessments and supporting records for at least three years, and notify the attorney general without unreasonable delay and no later than 90 days after discovering algorithmic discrimination. Compliance with the chapter creates a rebuttable presumption that a deployer used reasonable care in enforcement actions brought by the attorney general on or after July 1, 2027. The bill also requires deployers beginning July 1, 2026 to notify consumers and provide a plain‑language statement when a high‑risk AI system is used to make or substantially influence a consequential decision affecting a consumer, and it lists certain exemptions and exceptions (including an exemption for smaller deployers meeting specific conditions). Enforcement and procedural provisions treat violations as unfair or deceptive acts under the state consumer protection act, authorize the attorney general to bring actions in the name of the state or as parens patriae, require 45 days’ written notice before commencing an action, and allow a first violator a 60‑day cure period after receiving notice. The bill requires government agencies that make AI systems available to interact with consumers to disclose that the interaction involves an AI system, and it includes a range of listed nondisapplication clauses and exemptions (for example, certain federally regulated activities and research). The attorney general may prescribe forms and may designate a risk management framework. The bill also extends and expands an artificial intelligence task force and creates an artificial intelligence workplace advisory group administered and staffed by the attorney general, with required membership, subcommittee rules, posting of meeting summaries, and specified report deadlines (preliminary report due December 31, 2024; interim report due December 1, 2025; final report due July 1, 2027; workplace advisory group interim report due December 1, 2026 and final report due March 1, 2027). Those task force and workplace advisory group provisions expire June 30, 2028. Important details are incomplete in the provided text: Section 5 was truncated so the full content of the impact assessment requirements and section 5(6) are not visible here, section 9 (full enforcement text) is not included in the extracted facts, and the exact RCW chapter numbers to be added and any expiration dates mentioned in the bill header are not specified.

If enacted, companies that put AI systems into actual use to make important decisions about people will face new, concrete compliance work and costs: they will need to tell Washington consumers when such a system is being used (starting July 1, 2026), build and run an AI risk management program and do annual reviews of each high-risk system (starting July 1, 2027), prepare impact assessments when systems are deployed or substantially changed and keep those records for three years, and notify the state attorney general within 90 days if they discover the system caused unlawful bias. These steps are likely to increase operating expenses for affected businesses (staff time, expert consultants, recordkeeping, and possible system changes), raise the risk of enforcement actions by the attorney general (though compliance creates a rebuttable presumption of reasonable care and a notice-and-cure process for first violations), and give consumers more transparency about when AI is used in consequential decisions. The parties most affected are private-sector developers and deployers doing business in Washington, who will face new responsibilities and compliance costs; small deployers with fewer than 50 employees who meet specific conditions may avoid some requirements. The attorney general’s office will take on administration, enforcement, and staffing of an expanded AI task force and a workplace advisory group with set reporting deadlines, which could require additional resources. Important details are missing from the extracted text—full content of the impact assessment requirements, the exception referenced as section 5(6), some enforcement language, and exact statutory chapter numbers and expiration dates—so the precise scope of obligations and any budgetary impacts are not fully clear.