In a move that underscores just how quickly states are stepping in as federal oversight stalls, Washington’s Artificial Intelligence Task Force has released an interim report calling for sweeping AI regulation across sectors from healthcare to law enforcement. The Task Force’s recommendations are some of the most comprehensive in the country, laying out transparency mandates, rules for high-risk AI systems, and new limits on automated decision-making in areas like healthcare and employment.
The most striking part of the report is its blunt acknowledgment that the federal government has shown “little interest” in meaningful AI regulation. As a result, the Task Force argues, states must now take on the responsibility of protecting consumers, workers, and civil rights.
Key recommendations: The interim report outlines a number of core recommendations.
- Mandatory Transparency About AI Training Data
- The provenance of training datasets
- Data quality and quantity
- How data is processed to reduce errors or biases
This recommendation signals increasing scrutiny of the datasets behind underwriting models, scoring tools, conversational agents, and automated decision systems.
- Governance Requirements for “High-Risk” AI
- Mandatory adoption of recognized risk-management frameworks like NIST
- Public disclosure of risk-management practices
- Evaluation of whether certain high-risk uses should even be allowed
For organizations deploying AI decision tools in credit, collections, or consumer interactions, this raises questions about what regulators may soon classify as “high-risk.”
- Limits on AI in Healthcare Decisions
- AI cannot be used to deny, delay, or modify health services without a licensed clinician making the final call. While healthcare-specific, this signals a broad regulatory theme: AI cannot autonomously make adverse decisions affecting consumers.
- Public Disclosure of AI Use by Law Enforcement
- This includes attestations that AI-generated or AI-modified reports are accurate.
- Given how enforcement agencies nationwide use AI for fraud investigations and identity verification, this could influence regulatory expectations in other sectors.
- Investment in STEM + Broadband to Close AI Literacy Gaps
- The Task Force notes the digital divide directly impacts whether vulnerable communities can understand or challenge AI-driven decisions.
Why this matters:
- State-level AI laws are coming fast. Washington’s report mirrors emerging trends in Colorado, California, Texas, and New York.
- Expect stricter transparency requirements. Vendors supplying AI tools may soon need to disclose more detail about training data, risk controls, and governance.
- Human-in-the-loop requirements could expand beyond healthcare. Any AI that impacts a consumer’s rights or access to essential services may eventually need human oversight.
- Federal inaction means patchwork compliance. Agencies operating in multiple states may face divergent mandates.
Washington’s final report is due July 1, 2026, and is widely expected to influence legislative agendas nationwide. For now, the interim report makes one thing clear: states are preparing to regulate AI broadly, and industries touching sensitive consumer data should prepare accordingly.
