WA SB5984

Operators must provide clear disclosures that AI companion chatbots are artificially generated and not human, repeated at the start of each session and every 3 hours during use (every hour for minors)

AI companion chatbots are prohibited from claiming to be human, generating sexually explicit content for known minors, or using manipulative techniques like simulating abandonment feelings or prompting users to return for emotional support

Operators must implement protocols to detect suicidal ideation and self-harm expressions, provide automated crisis resource referrals, and publicly disclose the number of crisis referral notifications issued annually

The law applies to AI systems that retain user information across sessions, ask unprompted personal questions, and sustain ongoing personal dialogue—excluding customer service bots, in-game bots, and virtual assistants

Violations constitute unfair or deceptive practices under Washington's Consumer Protection Act (chapter 19.86 RCW), with an effective date of January 1, 2027