Washington state has enacted several targeted AI-related laws in recent years, focusing primarily on transparency for AI-generated content, safety for AI companion chatbots, protections against deepfakes, and ongoing study via a task force. There is no comprehensive “AI regulation” like some other states’ high-risk AI rules, but these measures address specific risks such as misinformation, harm to minors, and deception.2
Key Laws (as of April 2026)
1. HB 2225 (2026) – Regulation of AI Companion Chatbots
•Signed: March 24, 2026 (Chapter 168, Laws of 2026).
•Effective: January 1, 2027.
•Applies to: “Operators” of AI companion chatbots (systems designed for ongoing, human-like interactions/relationships, e.g., like advanced versions of ChatGPT or Claude for personal use). Excludes narrow-purpose tools like basic customer service bots.1
•Main requirements:
◦Safeguards for minors, including limits on manipulative, explicit, or harmful content.
◦Protocols to detect and respond to signs of self-harm, suicide ideation, or eating disorders (e.g., redirect to crisis hotlines).
◦Disclosures that the chatbot is AI (not human).
◦Enforcement by the Attorney General, plus a private right of action (users can sue). Violations treated as unfair/deceptive acts under the Consumer Protection Act.5
•This is described as the first-in-the-nation law specifically regulating AI chatbots for safety.4
2. HB 1170 (E2SHB 1170, 2026) – AI Content Provenance and Disclosures
•Signed: March 24, 2026 (Chapter 167, Laws of 2026).
•Effective: February 1, 2027.
•Applies to “Covered providers”: Entities producing generative AI systems with >1 million monthly users, publicly accessible in Washington for personal use (large companies like those behind major image/video/audio generators). Excludes government, B2B uses, video games, minor edits (e.g., cropping, brightness), and upscaling/noise reduction.36
•Main requirements:
◦Embed provenance data (e.g., watermarks or metadata, using standards like C2PA) in AI-created or “materially altered” images, video, or audio to allow users to verify origin/authenticity. Must be tamper-resistant where commercially/technically reasonable.
◦Provide detection tools for users.
◦Government agencies must disclose when interacting with AI systems.
•Enforcement: Attorney General only (Consumer Protection Act violations).2
3. Deepfake/Right of Publicity Protections (SSB 5886, 2026)
•Expands Washington’s existing right-of-publicity law to prohibit unauthorized AI-generated digital replicas (audio/video likenesses) of individuals. Signed in March 2026; effective June 11, 2026.6
AI Task Force (ESSB 5838, 2024)
•Established in 2024 and administered by the Attorney General’s Office.
•It studies AI uses, risks (including to equity, workforce, privacy, and ethics), benefits, and recommends future legislation/guidelines.
•Reports: Preliminary (2024), Interim (Dec 2025), Final (due July 1, 2026). It has held public meetings and produced a literature review.0
Other Context
•Governor’s Executive Order (24-01): Earlier guidance for state agencies on ethical/transparent use of generative AI.7
•Additional proposals (e.g., on high-risk AI discrimination, schools, labor bargaining) have been discussed but the above are the main enacted laws.
•Washington builds on prior narrower rules (e.g., facial recognition limits).13
These laws emphasize transparency, child safety, and consumer protection rather than broad bans or licensing. For full text, check the Washington Legislature site (
app.leg.wa.gov) or the AG’s AI Task Force page.