WeVote
Log In

Bill

Bill

HB 2667

Providing consumer protections for artificial intelligence systems.

2025-2026 Regular Session Introduced by Stephanie Barnard and 2 co-sponsors

HB 2667 requires AI system developers to disclose functionality and establish accountability standards, creating consumer protections against algorithmic harm in Washington.

First reading, referred to Technology, Economic Development, & Veterans.
0
WeVote Research Nonpartisan
Bill Summary · HB 2667

Legislative bill overview

HB 2667 establishes consumer protection requirements for artificial intelligence systems used in Washington state. The bill aims to create disclosure standards, liability frameworks, and safeguards for AI-driven decision-making that affects consumers. This marks an early legislative effort to regulate AI before deployment becomes more widespread.

Why is this important

As AI systems increasingly influence consumer outcomes—from credit decisions to hiring to content recommendations—regulatory gaps create potential for discriminatory harm, data misuse, and consumer confusion. This bill addresses whether companies deploying AI should be held accountable for algorithmic failures and whether consumers deserve transparency about AI use in decisions affecting them. The approach Washington takes could influence how other states develop AI oversight.

Potential points of contention

  • Business compliance costs: Requiring AI transparency and testing may create expensive compliance burdens, particularly for small tech companies, potentially slowing innovation or increasing consumer service costs
  • Definition and scope ambiguity: "Artificial intelligence systems" can mean different things; overly broad definitions might regulate basic automated systems while overly narrow ones miss emerging risks
  • Liability allocation: Determining whether liability falls on AI developers, deploying companies, or both creates uncertainty; excessive liability could discourage beneficial AI development while insufficient liability leaves consumers unprotected
  • Technical feasibility: Some AI systems (like deep learning models) are difficult to explain; mandating transparency may be technically impossible without fundamentally changing how certain systems work

Compiled from official sources — confirm details with the bill’s official record.

Sign in to ask a question.