SB 6120 creates a new chapter in Washington law regulating “high-risk artificial intelligence systems” used in the state. The bill defines a high-risk AI system as one intended to autonomously make or substantially factor into consequential decisions affecting areas such as employment, housing, education, lending, insurance, healthcare, legal services, parole/probation, and other significant consumer outcomes. It also defines related terms such as algorithmic discrimination, developers, deployers, generative AI, synthetic content, facial recognition, and general-purpose AI models, and it excludes a range of tools and uses from the high-risk category, including many standard software products, research/prototyping, and certain government, insurance, telemedicine, and federal-use systems.
The bill imposes duties on both developers and deployers of high-risk AI systems. Developers must use reasonable care to protect consumers from known or reasonably foreseeable risks of algorithmic discrimination, provide extensive documentation and disclosures to deployers, and identify synthetic content generated by certain systems. Deployers must adopt risk management policies, complete impact assessments before initial deployment and significant updates, disclose to consumers when they are interacting with AI, explain consequential decisions in certain adverse cases, and maintain records. The bill also creates a private right of action, allowing civil suits for violations, with possible injunctions and attorney fees, while providing a cure period as an affirmative defense. It states that the chapter is remedial, to be construed broadly for consumer protection, and takes effect January 1.
The overall sentiment reflected by the bill text is strongly consumer-protective and precautionary. The legislation is framed as a response to algorithmic discrimination and a desire for transparency, accountability, and risk management in AI systems that affect important life opportunities. The absence of committee transcripts or recorded votes means there is no available public debate in the provided materials to show support or opposition from specific legislators or stakeholders.
The main points of contention likely center on the breadth of the compliance obligations and the scope of the exemptions. Developers and deployers must produce detailed documentation, conduct impact assessments, maintain records, and provide consumer disclosures, which could be viewed as burdensome by AI vendors, employers, lenders, insurers, and other regulated entities. At the same time, the bill includes numerous carve-outs for federal systems, financial institutions under existing regulators, insurers, sandbox environments, telemedicine, and certain research or security-related uses, suggesting an effort to balance regulation with operational flexibility. Another likely issue is the bill’s treatment of trade secrets and proprietary information, which it protects from disclosure, and its potential interaction with First Amendment and other legal rights, which the bill expressly says it should not impair.
SB 6120 would add a new chapter to Title 19 RCW establishing statewide rules for high-risk AI systems. It would create new statutory duties for AI developers and deployers, require risk management and impact assessment processes, mandate consumer disclosures, and authorize private civil enforcement. The bill would also affect how AI is used in consequential decision-making across employment, housing, lending, insurance, healthcare, education, and related sectors, while preserving a broad set of exemptions and limiting disclosure obligations where trade secrets or other protected information are involved.
The bill appears to have a generally supportive, consumer-protection-oriented posture based on its text, which emphasizes preventing algorithmic discrimination and ensuring transparency and accountability. Because no committee transcripts or vote history were provided, there is no recorded legislative debate or roll-call evidence in the materials to indicate partisan division, amendments, or formal opposition. The framing suggests the bill is intended as a remedial measure rather than a permissive or deregulatory one.
Likely areas of contention include the compliance burden on AI developers and deployers, especially the requirements to document systems, conduct impact assessments, maintain records, and provide detailed consumer notices. Another likely dispute is the breadth of the bill’s definitions of high-risk AI and consequential decisions, which could sweep in many commercial uses, versus the extensive exemptions for federal systems, insurers, financial institutions, telemedicine, sandbox testing, and certain security or research activities. Stakeholders may also contest the private right of action, the cure period, and how the bill balances transparency with protection of trade secrets and proprietary information.