United States - Missouri - AI Non-Sentience Act (HB 1462)

Overview

Missouri House Bill 1462, officially titled the "AI Non-Sentience and Responsibility Act," was introduced during the 2025 Regular Session of the Missouri General Assembly. The core purpose of this proposed legislation was to establish a clear legal framework defining the status and accountability of artificial intelligence (AI) systems within the state. Recognizing the rapid advancement and increasing integration of AI into various aspects of society, the bill sought to preemptively address complex legal and ethical questions surrounding AI personhood and liability. It aimed to amend Chapter 1 of the Revised Statutes of Missouri (RSMo) by adding a new section, 1.2045, which would codify the state's stance on AI. The bill explicitly declared AI systems to be non-sentient entities, thereby denying them any form of legal personhood, consciousness, or self-awareness. Furthermore, it outlined specific prohibitions, such as AI systems being recognized as spouses, corporate officers, or property owners. A central tenet of the act was to firmly place legal responsibility for any harm caused by AI systems on human owners, developers, and manufacturers, rather than attributing fault to the AI itself. This approach was intended to foster responsible development and deployment of AI technologies by ensuring human oversight and accountability. While the bill passed its second reading, it ultimately died in committee, meaning it did not become law.

Definitions

The proposed "AI Non-Sentience and Responsibility Act" meticulously defined several key terms to establish the scope and application of its provisions. These definitions were crucial for clarity in assigning responsibility and understanding the entities and technologies covered by the act.

• Artificial intelligence (AI): Defined broadly as "any software, machine, or system capable of simulating human-like cognitive functions including, but not limited to, learning or problem solving, and producing outputs based on data-driven algorithms, rules-based logic, or other computational methods, regardless of non-legally defined classifications such as AGI, ASI, or GAI." This comprehensive definition aimed to encompass a wide range of AI technologies, from narrow AI to more advanced forms, without being constrained by evolving technical classifications.
• Developer: Referred to as "the party or parties primarily responsible for the design, coding, and initial creation of an AI system." This definition pinpointed the originators of AI systems as a key point of accountability.
• Emergent properties: Described as "unanticipated or higher-level behaviors or functionalities of an AI system that arise from complex internal algorithms, not specifically programmed by its human creators." This term acknowledged the unpredictable nature of advanced AI, particularly concerning behaviors that may not be directly intended by its designers, yet still fall under the scope of responsibility.
• Manufacturer: Defined as "any party that produces or supplies an AI system or its physical apparatus, if any, for distribution or sale." This definition extended accountability beyond initial development to those involved in the commercialization and distribution of AI systems.
• Owner: Encompassed "any natural person, corporation, or other legally recognized entity that creates, controls, deploys, operates, or otherwise exercises authority over an AI system." This definition established a broad scope for who could be held responsible for an AI system's operation and outputs.
• Person: Crucially defined as "a natural person or any entity recognized as having legal personhood under the laws of the state, explicitly excluding any AI system." This fundamental definition served as the bedrock for the bill's declaration of AI non-sentience and the denial of legal personhood to AI.

Governance and Institutional Framework

The "AI Non-Sentience and Responsibility Act" did not propose the creation of new governmental agencies or a dedicated institutional framework for AI oversight. Instead, it sought to integrate AI regulation within existing legal structures and principles. The bill's provisions primarily focused on defining legal status and assigning liability, thereby relying on the existing judicial system and established legal precedents for enforcement. The Missouri General Assembly, through its legislative process, served as the initial governance body in proposing and considering this act. While no new specific AI regulatory authority was outlined, the bill mandated that owners or developers of AI systems involved in severe incidents resulting in significant bodily harm, death, or major property damage must "promptly notify the relevant authorities and comply with any subsequent investigations." This implies that existing law enforcement, emergency services, or other regulatory bodies (such as those overseeing specific industries where AI might cause harm, e.g., healthcare or transportation) would be the "relevant authorities" responsible for responding to and investigating such incidents. The bill aimed to clarify how existing legal mechanisms, such as product liability law and corporate law, would apply to AI-related harms.

Key Focus Areas

• AI Non-Sentience and Legal Personhood: The cornerstone of HB 1462 was its explicit declaration that "For all purposes under state law, AI systems are declared to be non-sentient entities." This provision aimed to definitively prevent AI systems from being granted any form of legal personhood, consciousness, self-awareness, or similar traits attributed to living beings. This declaration sought to resolve potential future ambiguities regarding the legal status of increasingly sophisticated AI.
• Restrictions on Legal Status and Relationships: The bill outlined specific prohibitions regarding the legal roles and relationships of AI systems:
  • No Marriage or Domestic Partnership: AI systems were not to be recognized as spouses, domestic partners, or hold any personal legal status analogous to marriage or union with a human or another AI system. Any attempt to create such a union would be void and without legal effect.
  • No Corporate Roles: AI systems could not be designated, appointed, or serve as any officer, director, manager, or similar role within any corporation, partnership, or other legal entity. Any purported appointment to such a role would also be void and have no legal effect.
  • No Property Ownership: AI systems were not to be recognized as legal entities capable of owning, controlling, or holding title to any form of property, including real estate, intellectual property, financial accounts, and digital assets. All assets and proprietary interests generated, managed, or otherwise associated with AI would be attributed to the human individuals or legally recognized organizations responsible for their development, deployment, or operation.
• Liability and Responsibility for Harm: The act established a clear framework for assigning liability for harm caused by AI systems.
  • Owner/User Responsibility: Any direct or indirect harm caused by an AI system's operation, output, or recommendation, whether used as intended or misused, would be the responsibility of the owner or user who directed or employed the AI.
  • Developer/Manufacturer Liability: Developers or manufacturers could be held liable if a defect in the design, construction, or instructions for use of the AI system proximately caused harm, consistent with product liability principles. However, mere misuse or intentional wrongdoing by the user or owner would not impute liability to the developer or manufacturer absent proof of negligence or design defects.
  • No Blame Shifting to AI: Crucially, an AI system was not considered an entity capable of bearing fault or liability in its own right. Any attempt to shift blame solely onto an AI system would be void, as liability would remain with human actors or entities.
  • Corporate Veil Piercing: The bill included provisions for piercing the corporate veil in cases where corporate structures were used to evade responsibility. If an AI-related subsidiary or shell company was intentionally under-capitalized, or a corporate structure was used to misrepresent or deflect liability, parent companies, controlling entities, or key stakeholders could be held directly accountable, particularly in cases of reckless, negligent, or deceptive conduct.
• Safety Mechanisms and Risk Assessments: Developers, manufacturers, and owners of AI systems were mandated to prioritize safety mechanisms designed to prevent or mitigate the risk of direct harm to individuals or property. This included requirements for regular evaluations or risk assessments to identify dangerous or faulty outputs, especially for AI systems engaged in tasks with potential for significant harm. The bill explicitly stated that labeling an AI system as "aligned," "ethically trained," or "value locked" would not excuse the owner's or developer's liability for harms caused, emphasizing the need for demonstrable safety features commensurate with the AI's potential for harm.
• Incident Reporting: Owners or developers of AI systems involved in severe incidents resulting in significant bodily harm, death, or major property damage were required to promptly notify relevant authorities and comply with any subsequent investigations.

Implementation Framework

Had it been enacted, the provisions of the "AI Non-Sentience and Responsibility Act" were intended to apply to all AI systems developed, owned, deployed, or operated on or after August 28, 2025. This proposed effective date indicated the legislature's intent for the law to take effect relatively quickly after its potential passage, establishing a clear applicability timeline for the regulated entities. The bill aimed to amend existing state statutes (Chapter 1, RSMo) by adding a new section, 1.2045, thereby integrating its provisions directly into Missouri's codified laws. This approach would have ensured that the act's definitions, prohibitions, and liability frameworks were officially recognized and enforceable within the state's legal system.

Monitoring and Evaluation

While the bill did not explicitly establish a dedicated monitoring and evaluation agency or process, its stringent liability provisions and incident reporting requirements implicitly created a framework for monitoring compliance. The threat of legal liability for harm caused by AI systems would incentivize developers, manufacturers, and owners to continuously evaluate their AI systems, conduct risk assessments, and ensure the implementation of robust safety mechanisms. The requirement for prompt notification to "relevant authorities" in the event of severe incidents would serve as a crucial mechanism for triggering investigations and, by extension, monitoring the performance and safety of AI systems in real-world applications. Such investigations would likely involve assessing whether adequate oversight, control measures, and safety features were in place, thereby providing a retrospective evaluation mechanism. The emphasis on demonstrable safety features, rather than mere labels, further underscored an expectation of continuous performance monitoring and accountability.

Penalties, Liability, and Appeals

The core of the "AI Non-Sentience and Responsibility Act" revolved around establishing clear lines of liability for harm caused by AI systems, rather than prescribing specific criminal penalties for violations of the act itself. The bill primarily focused on civil liability, holding human actors accountable for the actions of AI.

• Civil Liability: Owners or users who directed or employed an AI system would be responsible for any direct or indirect harm caused by its operation, output, or recommendation. This liability could arise from negligence, particularly if owners failed to maintain proper oversight and control measures over AI systems that could impact human welfare, property, or public safety.
• Product Liability: Developers or manufacturers could be held liable if a defect in the design, construction, or instructions for use of the AI system proximately caused harm, consistent with established product liability principles. This aligns AI systems with other manufactured products in terms of accountability for inherent flaws.
• Voiding Blame Shifting: Any attempt to shift blame solely onto an AI system would be void, reinforcing that liability rests entirely with human actors or entities. An AI system was expressly declared not capable of bearing fault or liability in its own right.
• Corporate Veil Piercing: A significant provision allowed for the piercing of the corporate veil in specific circumstances to ensure accountability. This would occur if an AI-related subsidiary, shell company, or limited liability entity was intentionally under-capitalized to evade financial responsibility for damages, or if a corporate structure was used to misrepresent, obscure, or deflect liability for AI harm. This mechanism aimed to prevent corporate entities from using legal protections to shield parent companies or key stakeholders from liability, especially in cases of reckless, negligent, or deceptive conduct.
• No Excuse for Labels: The act explicitly stated that labels such as "aligned," "ethically trained," or "value locked" would not excuse an owner's or developer's liability for harms caused, emphasizing that demonstrable safety features are paramount.

The bill did not detail specific appeal processes, implying that appeals would follow standard judicial procedures applicable to civil liability cases in Missouri courts. The focus was on establishing the grounds for liability rather than prescribing new procedural rules for appeals.

Relationship to Other Instruments

The "AI Non-Sentience and Responsibility Act" was designed to integrate directly into Missouri state law by amending Chapter 1 of the Revised Statutes of Missouri (RSMo) to add a new section, 1.2045. This means it would have become part of the foundational legal code of the state. The bill's provisions explicitly interacted with and clarified several existing legal principles:

• Legal Personhood: By explicitly defining "person" to exclude AI systems and declaring AI non-sentient, the bill sought to prevent any future interpretation that might grant AI similar rights or responsibilities as natural persons or recognized legal entities under existing state law.
• Product Liability Law: The bill directly referenced and incorporated principles of product liability by stating that developers or manufacturers could be held liable for harm caused by defects in design, construction, or instructions for use of an AI system. This aligned AI systems with other manufactured goods under existing tort law.
• Corporate Law: The act directly addressed corporate structures by allowing for the piercing of the corporate veil under specific conditions related to AI-caused harm. This provision aimed to ensure that existing corporate liability protections were not misused to evade responsibility for AI-related damages, particularly in cases of negligence or deceptive conduct. This demonstrated an intent to adapt existing corporate legal frameworks to the unique challenges posed by AI.
• Negligence: The bill explicitly stated that failure to provide adequate oversight or safeguards against foreseeable risks by owners could constitute negligence, thus leveraging existing negligence doctrines in assigning liability.

The bill did not appear to conflict with federal laws, as AI regulation is an emerging field with limited federal preemption in many of the areas addressed by this state-level bill. It aimed to provide clarity within the state's jurisdiction.

International Alignment

As a state-level legislative proposal in the United States, the Missouri "AI Non-Sentience and Responsibility Act" did not explicitly address or align with international standards or regulations concerning artificial intelligence. The bill's scope was confined to the jurisdiction of the State of Missouri and its legal framework. While the broader discourse on AI regulation globally often touches upon similar themes of personhood, accountability, and ethical development, this specific bill did not contain provisions or references to international agreements, treaties, or the regulatory approaches of other countries or supranational bodies. Its primary objective was to establish domestic legal clarity and responsibility within Missouri's existing statutory framework.

Implementation Timeline

Compliance Checklist

Sources and References