March 27, 2026

TOP LINE

Washington State's HB1170 — requiring disclosure when content has been developed or modified by AI — has been delivered to the governor, making it the most immediate decision point in today's landscape. A signature would add Washington to a growing roster of states with affirmative AI labeling mandates, compounding compliance obligations for multi-jurisdictional content platforms. Oregon has separately passed SB1546 governing AI companion products, and Wyoming has enacted deepfake protections for minors, each adding discrete statutory perimeters that product and compliance teams must now map. Today's briefing reflects predominantly US state legislative activity; EU and international coverage was limited.

KEY DEVELOPMENTS

Washington HB1170 advances to governor. The bill, which mandates disclosure when AI has developed or modified content delivered to users, now awaits executive action — the next concrete decision point. This fits a pattern of state-level AI transparency mandates that, cumulatively, create a patchwork of disclosure triggers varying by jurisdiction, audience, and content type. Product counsel at platforms distributing AI-generated or AI-modified content should assess whether current disclosure mechanisms satisfy the bill's requirements before a signature lands. Watch level: ACT NOW (content platforms, digital media operators, product counsel) — contingent on signature, but drafting should begin now.

Oregon passes AI companion bill; Wyoming enacts minor deepfake law. Oregon's SB1546 governing AI companion products has been signed by the House Speaker, and Wyoming's HB0102 criminalizing AI-generated exploitative imagery of minors has been assigned a chapter number, confirming enactment. Both represent enacted law, not pending proposals. Oregon's scope — consumer protection and psychological safety in emotionally-designed AI systems — raises questions about what disclosure, design, or operational requirements apply to relationship or social AI products; final text review is needed. Wyoming's law reinforces a clear state-level enforcement trend targeting AI-generated CSAM, consistent with earlier enactments in other states. Watch level: ACT NOW (AI companion developers, image-generation platform operators, content moderation teams)

Virginia passes two bills institutionalizing independent AI verification. SB384 passed 99-0 and HB797 passed 84-14, both establishing frameworks for independent AI verification organizations. The near-unanimous votes signal political durability for this structural approach. The practical question is whether these frameworks will require accreditation standards, conflict-of-interest rules, or specific audit scopes — details that will determine whether existing third-party assessors qualify or whether new organizational categories emerge. Compliance teams advising AI developers in Virginia should monitor enrollment and any implementing regulations. Watch level: PREPARE (AI developers subject to Virginia oversight, AI auditors and assessors, governance leads)

New York advances training data transparency bills. Both S06955 and A06578 — companion Senate and Assembly versions of the Artificial Intelligence Training Data Transparency Act — have advanced to third reading. Advancement of both chambers' versions simultaneously suggests coordinated legislative momentum. If enacted, the bills would require public disclosure of training data sourcing on developer websites, creating a novel supply-chain accountability obligation with no current federal analog. Developers operating at scale should begin inventorying training data provenance now, as the disclosure requirements could be difficult to satisfy retroactively. Watch level: PREPARE (generative AI developers, foundation model providers, AI product counsel)

Colorado advances two AI bills; Washington's high-risk AI bill shelved. Colorado's HB1139 on AI in health care passed the House Second Reading and HB1263 on conversational AI operators has been introduced and assigned to committee — continuing Colorado's posture as an active state-level AI regulator. Separately, Washington's HB2157 on high-risk AI systems has been placed in the House Rules 'X' file, effectively ending its prospects this session. The Washington stall suggests that broad horizontal AI risk frameworks continue to face procedural resistance even in active legislative states, while sector-specific and operator-specific bills move more readily. Health AI developers and conversational AI operators in Colorado should prioritize the two advancing bills. Watch level: PREPARE (Colorado health AI developers, conversational AI operators) / MONITOR (Washington high-risk AI stakeholders)

FORWARD LOOK

The Washington governor's decision on HB1170 is the nearest hard deadline in this cycle; no public timeline has been confirmed, but the delivery to the executive office starts the clock. Utah's HB0450 privacy amendments and HB0320 on the AI Policy Office are both awaiting gubernatorial action and may resolve in the near term. New York's dual training data transparency bills reaching third reading simultaneously warrants close tracking — if the Senate and Assembly versions are reconciled, a floor vote may follow before session close, though timing remains uncertain. Virginia's AI verification framework bills moving toward enrollment will likely require monitoring for implementing regulations that define accreditation standards.