SB 6284 Washington Senate · 2025-2026 Regular Session

Providing consumer protections for artificial intelligence systems.

Washington State's SB 6284 creates new consumer protections for artificial intelligence systems that make consequential decisions affecting people's lives. The bill requires businesses deploying "high-risk" AI systems (like those used for housing, loans, employment, or healthcare access) to conduct bias audits before use and disclose when AI influences such decisions. It defines "algorithmic discrimination" as AI causing unlawful bias and sets a risk-based regulatory framework for developers and deployers. The law applies to Washington residents acting as consumers, not in business or employment contexts, and aims to prevent unfair outcomes while supporting innovation.

Bill status in committee 1 of 4 stages cleared

Introduction

Jan 2026

Committee Review

Floor Vote

Governor

Introduced Jan 22, 2026 Last action Feb 6, 2026

Maddy AI version diff · 1 comparison

What changed between versions

Bill → Substitute Bill · 8 edits

MODERATE

The bill was renumbered from S-3753.4 to S-4569.2 and updated to reflect its status as a Substitute Bill with a new reading date. The core policy changes redefine 'high-risk' AI systems to require meaningful human oversight, expand exemptions for small businesses and healthcare providers, and shift enforcement authority to the Attorney General with a 45-day cure period for violations.

Scope change

The bill's scope was narrowed by adding specific exemptions for small businesses (fewer than 50 employees), private clubs, and certain healthcare and insurance entities, while simultaneously tightening the definition of 'high-risk' systems to exclude those with meaningful human consideration.

DEFINITION

The definition of 'high-risk artificial intelligence system' was changed to require 'meaningful human consideration' to avoid being classified as high-risk, and the definition of 'substantial factor' was updated to focus on factors generated by AI.

New definitions were added for 'meaningful human consideration' and 'consumer' (limiting it to individual/household contexts, excluding commercial use).

ELIGIBILITY

Exemptions were expanded to include developers with fewer than 50 employees and deployers who do not use their own data to train the system.

REQUIREMENT

Risk management policy requirements were restructured to apply separately to both developers and deployers, with specific criteria for each.

New requirements were added for insurers and financial institutions, including specific exemptions for those regulated by the Insurance Commissioner.

The disclosure requirement for AI interactions was updated to remove the specific prohibition on 'high-risk' systems, applying broadly to all AI interactions with consumers.

ENFORCEMENT

Enforcement was centralized under the Attorney General with a new provision allowing a 60-day cure period for the first violation.

TIMELINE

The expiration date for the AI task force was extended from June 30, 2027, to June 30, 2028.

Floor votes

How they voted

No floor votes recorded yet.

Full legislative history

Actions timeline

Total actions

Key actions

Committee

Feb 6, 2026

Upper · Passed

Public hearing in the Senate Committee on Ways & Means at 1:30 PM.

upper

Feb 4, 2026

Committee

Referred to Ways & Means.

upper

Feb 3, 2026

Upper · Passed

Minority; without recommendation.

upper

Feb 3, 2026

Upper · Passed

Minority; do not pass.

upper

Feb 3, 2026

Committee

And refer to Ways & Means.

upper

Feb 3, 2026

Upper · Passed

Executive action taken in the Senate Committee on Environment, Energy & Technology at 1:30 PM.

upper

Jan 27, 2026

Upper · Passed

Public hearing in the Senate Committee on Environment, Energy & Technology at 1:30 PM.

upper

1 primary · 8 co-sponsors