SB 1012: Creates new provisions relating to artificial intelligence - Missouri

4/30/2026 - SA 1 to SS#2 for SCS S offered & adopted (Carter)--(5687S16.04S) 4/30/2026 - Adopted → 4/30/2026 - SA 2 to SS#2 for SCS S offered & adopted (Burger)--(5687S16.01S) 4/30/2026 - Adopted · 3 edits

MINOR

The amendment changes the section of the bill being modified from Section 573.120 (Pages 15-16) to Section 1.2045 (Page 3). It replaces a provision that would have deleted an existing section with a new requirement stating that AI systems must not identify themselves as male or female.

Scope change

The scope of the amendment shifted from removing content to adding a specific gender-neutrality requirement for AI systems.

REQUIREMENT

Added a mandate requiring AI systems to refrain from identifying themselves as male or female.

Removed a provision that would have deleted Section 573.120 from the bill.

TECHNICAL

Updated the amendment instructions to target Section 1.2045 on Page 3 instead of Section 573.120 on Pages 15-16.

4/30/2026 - SS#2 for SCS S offered (Nicola)--(5687S.16F) 4/30/2026 - Adopted, as amended → 4/30/2026 - SA 1 to SS#2 for SCS S offered & adopted (Carter)--(5687S16.04S) 4/30/2026 - Adopted · 4 edits

MODERATE

The bill was amended to remove the entire text of the 'AI Non-Sentience and Responsibility Act' and its related definitions, effectively deleting the legislation that would have legally declared AI systems as non-sentient entities and restricted their ability to hold property or serve in leadership roles. This change reverses the bill's original intent to establish a regulatory framework for artificial intelligence, leaving the specific provisions regarding AI liability and status out of the final version.

Scope change

The bill's scope was completely removed; it no longer amends chapters 1, 130, or 573 of the Revised Statutes of Missouri to include new sections on artificial intelligence.

DEFINITION

All definitions for terms such as 'Artificial intelligence', 'Developer', 'End user', and 'Operator' were deleted from the bill.

REQUIREMENT

The requirement for government entities to deny AI systems legal personhood and the prohibition on AI systems owning property or holding executive positions were removed.

Provisions requiring clear notice when consumers interact with AI systems and rules regarding professional liability when using AI were deleted.

ENFORCEMENT

The severability clause and penalty provisions associated with the AI act were removed.

3/30/2026 - SA 4 to SS for SCS S offered & adopted (Gregory-15)--(5687S10.08S) 4/30/2026 - Moot → 4/30/2026 - SS#2 for SCS S offered (Nicola)--(5687S.16F) 4/30/2026 - Adopted, as amended · 4 edits

MODERATE

The bill was converted from a Senate Amendment to a Senate Substitute, indicating a complete rewrite of the text rather than a minor edit. The new version establishes the 'AI Non-Sentience and Responsibility Act,' which legally declares AI systems as non-sentient entities without personhood, property rights, or the ability to hold leadership positions. It also mandates that licensed professionals retain final authority over their decisions and remain fully liable for any errors made when using AI tools.

Scope change

The bill's scope expanded from a single provision regarding settlement confidentiality to a comprehensive four-section act covering AI definitions, legal status, professional liability, and consumer notice requirements.

DEFINITION

Added specific legal definitions for AI systems, developers, operators, owners, and stakeholders.

ELIGIBILITY

Established that AI systems are legally non-sentient and cannot be granted personhood, marry, own property, or serve as corporate officers.

REQUIREMENT

Required businesses to provide clear notice to consumers when interacting with AI systems.

ENFORCEMENT

Mandated that licensed professionals maintain independent judgment and bear full liability for mistakes made while using AI.

3/30/2026 - SA 3 to SS for SCS S offered & adopted (Beck)--(5687S10.01S) 4/30/2026 - Moot → 3/30/2026 - SA 4 to SS for SCS S offered & adopted (Gregory-15)--(5687S10.08S) 4/30/2026 - Moot · 4 edits

MODERATE

The amendment completely replaces a section that restricted healthcare boards from granting regulatory exceptions to AI companies with a new section addressing settlement agreements. The removed text would have barred medical boards from waiving rules for AI developers, while the added text prohibits confidentiality clauses in legal settlements to ensure transparency in court cases.

Scope change

The bill's scope shifted from regulating AI-specific exemptions for healthcare boards to regulating disclosure requirements in legal settlements and consent decrees.

REQUIREMENT

Deleted a provision that forbade healthcare boards from granting regulatory mitigation or waivers to AI developers.

ENFORCEMENT

Added a new requirement that settlement agreements cannot include terms prohibiting the disclosure of facts, allegations, or evidence.

Added a provision stating that any confidentiality clause violating the new disclosure rule is void and unenforceable.

SCOPE

Changed the bill's focus from AI regulation to legal settlement transparency.

3/30/2026 - SA 2 to SS for SCS S offered & adopted (Hudson)--(5687S10.06S) 4/30/2026 - Moot → 3/30/2026 - SA 3 to SS for SCS S offered & adopted (Beck)--(5687S10.01S) 4/30/2026 - Moot · 4 edits

MODERATE

The bill was completely rewritten to shift focus from regulating consumer chatbots to regulating artificial intelligence used by medical professionals. The new version removes all provisions regarding age verification, user accounts, and restrictions on chatbot content, replacing them with a mandate that medical licensing boards cannot grant regulatory waivers to businesses developing AI for dispensing or prescribing medication.

Scope change

The bill's scope changed from regulating public-facing AI chatbots and protecting minors to regulating the regulatory flexibility of medical licensing boards regarding AI in healthcare.

REQUIREMENT

All requirements for age verification, user account creation, and restrictions on chatbot content (such as prohibitions on soliciting minors or discussing suicide) were removed.

DEFINITION

Definitions for 'companion chatbot', 'covered entity', 'minor', and 'sexually explicit conduct' were removed.

A new definition of 'artificial intelligence' was added, specifically defining it as a system trained on data that simulates human conversation with limited oversight.

ELIGIBILITY

A new requirement was added stating that medical licensing boards cannot grant regulatory mitigation or waivers to entities developing AI for dispensing or prescribing medications.

3/30/2026 - SA 1 to SS for SCS S offered & adopted (Moon)--(5687S10.05S) 4/30/2026 - Moot → 3/30/2026 - SA 2 to SS for SCS S offered & adopted (Hudson)--(5687S10.06S) 4/30/2026 - Moot · 4 edits

MODERATE

The bill was amended to add a new section establishing the 'Guidelines for User Age-Verification and Responsible Dialogue Act of 2026' (GUARD Act). This change introduces strict regulations for 'companion chatbots,' requiring mandatory age verification processes and prohibiting these AI systems from encouraging minors to engage in sexually explicit conduct or self-harm.

Scope change

The bill now specifically targets artificial intelligence systems designed to meet emotional or social needs, excluding standard customer service bots, video game features, and internal corporate tools.

DEFINITION

Defined 'companion chatbot' as AI systems designed for emotional/social needs, while explicitly exempting customer service bots, video game features, and internal business tools.

REQUIREMENT

Mandated that all users must create an account with verifiable age data before accessing a companion chatbot, with specific rules for existing accounts and new sign-ups.

Prohibited simple methods like asking for birth dates for age verification, requiring instead authenticated measures such as government ID.

ENFORCEMENT

Established criminal penalties, including fines of up to $100,000 per offense, for designing or operating chatbots that encourage minors toward self-harm or sexually explicit content.

5687S.16P - Perfected → 3/30/2026 - SA 1 to SS for SCS S offered & adopted (Moon)--(5687S10.05S) 4/30/2026 - Moot · 3 edits

MINOR

The bill was transitioned from a perfected Senate Substitute to a Senate Amendment, changing the document header and page numbering. The substantive text of the 'AI Non-Sentience and Responsibility Act' remains unchanged, maintaining the core provisions that AI systems are non-sentient, cannot hold legal personhood, and cannot serve in executive roles.

Scope change

The scope of the bill's applicability and the specific legal provisions remain unchanged; only the legislative document metadata and formatting were updated.

TECHNICAL

The document header was changed from 'Senate Substitute No. 2' to 'Senate Amendment No. ___', indicating a procedural shift in the bill's status.

Page numbering and section headers were reorganized to reflect the new amendment format.

A new line was inserted regarding federal law precedence, though this appears to be a formatting artifact within the amendment block rather than a substantive policy addition.

5687S.02C - Senate Committee Substitute → 5687S.16P - Perfected · 5 edits

MODERATE

The bill was amended to remove Chapter 573 from its scope and updated the list of amended chapters to include a new section in Chapter 1. The text was significantly rewritten to modernize definitions of artificial intelligence and add specific protections for licensed professionals, ensuring they retain final decision-making authority when using AI tools. It also clarifies that AI systems cannot identify as male or female and strengthens liability rules to ensure human actors remain responsible for AI outcomes.

Scope change

The bill's scope was narrowed by removing Chapter 573 from the list of chapters being amended, and the new text in Chapter 1 now includes a specific exemption for licensed professionals who use AI to assist in their work.

SCOPE

Chapter 573 was removed from the list of chapters being amended, and a new section was added to Chapter 1.

DEFINITION

The definition of 'Artificial intelligence' was completely rewritten to use modern terminology, and a new definition for 'End user' was added.

REQUIREMENT

New requirements mandate that licensed professionals must exercise independent judgment over AI recommendations and bear responsibility for mistakes made when using AI in their regulated services.

Liability rules were updated to explicitly state that owners or operators interacting with consumers must provide clear notice that they are using an AI system.

ADDED

New provisions state that an AI system cannot identify itself as male or female.

5687S.01I - Introduced → 5687S.02C - Senate Committee Substitute · 5 edits

MODERATE

The bill was amended to expand its scope from just chapters 130 and 573 to include chapter 1, and the number of new sections increased from two to three. The most significant changes are the addition of a new 'AI Non-Sentience and Responsibility Act' which explicitly declares AI systems as non-sentient entities and voids any legal attempts to grant them personhood, marriage rights, or ownership. The bill also broadens the definition of regulated technology from 'generative AI' to a wider 'artificial intelligence' and introduces new liability rules holding developers and manufacturers responsible for AI-caused harm while protecting users from being blamed for AI errors.

Scope change

The bill's scope expanded from chapters 130 and 573 to include chapter 1, and the regulated technology definition was broadened from 'generative artificial intelligence' to 'artificial intelligence' to cover a wider range of systems.

DEFINITION

Added a new section (1.2045) creating the 'AI Non-Sentience and Responsibility Act' which explicitly states AI systems are not sentient beings and cannot hold legal personhood, marry, own property, or be held liable for their actions.

Expanded the definition of regulated technology from 'generative artificial intelligence' to 'artificial intelligence' to include any software or system capable of simulating human cognitive functions or making autonomous decisions.

Removed specific definitions and disclaimer requirements related to 'generative artificial intelligence' in political advertisements, as these were superseded by the broader new definitions and the new non-sentience section.

REQUIREMENT

Added new liability provisions stating that developers and manufacturers are responsible for harm caused by AI defects, while owners and operators are responsible for harm caused by their use of the system.

Added a requirement for owners and operators to maintain reasonable oversight and risk controls for AI systems that could impact human welfare or public safety.

Creates new provisions relating to artificial intelligence

What changed between versions

How they voted

Actions timeline

Sponsors