SHB 1168
In CommitteeHouse
Artificial intelligence info
Increasing transparency in artificial intelligence.
This status may be delayed. See Action History below for the latest updates.
How does a bill become law?
- Introduced: The bill is filed and assigned a number.
- Committee: A subject-matter committee holds hearings, takes public testimony, and decides whether to advance the bill.
- Floor Vote: The full chamber (House or Senate) debates and votes on the bill.
- Opposite Chamber: The bill repeats the committee and floor vote process in the other chamber.
- Governor: The Governor reviews the bill and decides whether to sign or veto it.
- Signed: The bill has been signed into law.
AI Analysis
This bill requires developers of public-facing generative AI systems to publicly disclose detailed information about the data used to train those systems, including sources, volume, and whether personal or copyrighted data were included. It also establishes enforcement by the attorney general with civil penalties for noncompliance.
- Requires developers of public-facing generative AI systems to post detailed documentation on their websites about the data used to train those systems, including sources, volume, types, and whether personal or copyrighted data were used.
- Mandates disclosure of whether synthetic data generation was used in development, and if so, the functional purpose of that data.
- Exempts AI systems used solely for security and integrity, aircraft operations, or national security/military/defense purposes (when only provided to federal entities) from the disclosure requirement.
- Authorizes the attorney general to enforce the law and impose civil penalties of $5,000 per violation, with each day of noncompliance treated as a separate violation.
- Defines key terms such as *generative artificial intelligence*, *aggregate consumer information*, *developer*, and *substantially modifies* to clarify the scope and application of the law.
Who is affected
- Developers of generative AI systems or services made available to the public in Washington — Must publicly disclose documentation about the data used to train their generative AI systems, including sources, types, volume, and whether personal or copyrighted data were used.
- Washington residents who use public-facing generative AI tools (e.g., chatbots, image generators) — Benefit from increased transparency about how AI systems they use are trained, helping them make more informed decisions about AI tools.
- AI developers operating in Washington who do not meet the transparency requirements — May face civil penalties of $5,000 per violation if they fail to comply with disclosure requirements.
- Federal agencies and developers of AI for national security, military, or aircraft operations — May rely on AI systems for security, aviation, or defense purposes and are exempt from the disclosure rules under certain conditions.
Pro/Con Analysis
Stronger case for benefits
Potential Benefits (5)
Requires developers to disclose whether personal or aggregate consumer information was used in training, empowering Washington residents to make informed choices about AI tools and reducing risks of unauthorized data use—especially important for vulnerable populations (e.g., low-income users, non-English speakers) who may be disproportionately targeted by biased or exploitative AI.
Rights & LibertiesPeopleRef: Sec. 2(1)(vii), Sec. 2(1)(viii)Mandates public disclosure of data sources and purpose, enabling educators, students, and researchers to evaluate bias, accuracy, and reliability of AI tools used in classrooms—supporting digital literacy and critical thinking in K–12 and higher education across Washington.
EducationPeopleRef: Sec. 2(1)(i), Sec. 2(1)(ii), Sec. 2(1)(v)Requires disclosure of synthetic data use and aggregate consumer information, helping clinicians and patients assess whether AI diagnostic or triage tools used in Washington health systems are based on representative or synthetic data—reducing risks of misdiagnosis or inequitable care.
HealthcarePeopleRef: Sec. 2(1)(xii), Sec. 2(1)(viii)Disclosing whether datasets include copyrighted material helps prevent unauthorized use of creators’ work, supporting Washington artists, writers, and journalists whose intellectual property may otherwise be exploited without consent or compensation.
Rights & LibertiesPeopleRef: Sec. 2(1)(v), Sec. 2(1)(vi)Enforcement authority given to the attorney general creates a baseline of accountability, deterring bad-faith data scraping and encouraging ethical AI development—potentially fostering trust in AI tools among Washington consumers and supporting a more responsible local tech ecosystem.
Business & EmploymentPeopleRef: Sec. 3
Potential Concerns (5)
Mandates disclosure of whether personal or aggregate consumer information was used in training, which may incentivize developers to avoid collecting or using such data altogether—even when lawfully and ethically sourced—leading to reduced functionality or accuracy of AI tools for Washington residents who rely on personalized services (e.g., health diagnostics, financial advice, education tools).
Rights & LibertiesPeopleRef: Sec. 2(1)(vii), Sec. 2(1)(viii)Requires developers to disclose whether datasets include copyrighted material or were purchased/licensed, increasing legal exposure for developers who may have used data under ambiguous licensing or fair use doctrines; this could deter innovation and lead to consolidation as smaller developers may lack legal resources to navigate complex copyright compliance.
Business & EmploymentLean peopleRef: Sec. 2(1)(v), Sec. 2(1)(vi)Imposes civil penalties of $5,000 per *day* of noncompliance, creating significant financial risk for small developers or startups that may misinterpret vague terms like “substantially modifies” or “aggregate consumer information”—potentially forcing closures or reducing AI product offerings in Washington.
Business & EmploymentPeopleRef: Sec. 3Exempts national security, military, and defense AI systems from disclosure, creating a transparency gap where Washington residents may be subject to AI-driven surveillance or decision-making (e.g., immigration enforcement, border security) without public oversight or accountability.
Public SafetyLean peopleRef: Sec. 2(2)(c)Requires disclosure of data collection timeframes and synthetic data use, which may expose trade secrets or competitive strategies, disproportionately harming small firms that rely on proprietary AI development methods—potentially chilling innovation in Washington’s emerging tech sector.
Business & EmploymentLean peopleRef: Sec. 2(1)(xi), Sec. 2(1)(xii)
Who Is Most Affected
Small-to-midsize AI startups and independent developers in Washington may face disproportionate compliance costs relative to revenue, potentially reducing innovation or forcing exit from the market—though larger firms can absorb costs more easily.
May benefit from increased trust in AI tools and better-informed usage, but could face reduced access to high-quality or personalized AI services if developers avoid using personal data to minimize legal risk.
Large tech firms (e.g., Microsoft, Amazon) with existing legal and compliance teams can more easily meet disclosure requirements, potentially consolidating market dominance while smaller competitors struggle—net effect is mixed but leans negative for competition.
Educational institutions and public agencies using AI for teaching, research, or service delivery may benefit from transparency but could be constrained if developers reduce data use to avoid disclosure—potentially limiting tool functionality.
Content creators (writers, artists, musicians) may gain leverage to negotiate licensing or opt-out of data use, but enforcement is limited to AG actions—individuals have no private right of action, limiting real-world impact.