Establishes the "AI Nonsentience and Responsibility Act"

SECOND REGULAR SESSION
HOUSE COMMITTEE SUBSTITUTE FOR
HOUSE BILL NOS. 1746 & 1769
103RD GENERAL ASSEMBL Y
3891H.04C JOSEPH ENGLER, Chief Clerk
AN ACT
T o amend chapter 1, RSMo, by adding thereto one new section relating to artificial
intelligence.
Be it enacted by the General Assembly of the state of Missouri, as follows:
Section A. Chapter 1, RSMo, is amended by adding thereto one new section, to be
2 known as section 1.2045, to read as follows:
1.2045. 1. The prov isions of this section shall be known and may be cited as the
2 "AI Nonsentience and Responsibility Act".
3 2. For purposes of this section, the following terms mean:
4 (1) "Artificial intelligence" or "AI", the same meaning as used in 15 U.S.C.
5 Section 9401(3);
6 (2) "Deployer", any natural person, corporation, or other legally r ecognized
7 entity that creat es, contr ols, deploys, operates, or otherwise exerci ses authority over an
8 AI system;
9 (3) "Developer", the party or parties primarily r esponsible for the design,
10 coding, and initial cr eation of an AI system, unless the AI system is released fr ee of
11 charge using an Open Sour ce Initiative appr oved license;
12 (4) "Emergent prop erties", any unpr ogrammed or adaptive behavior of an AI
13 system that appears to mimic human sentience;
14 (5) "End-user", a person or agent of a person who interacts with an AI system.
15 The end-user does not have control over the AI system but r elies on it for specific tasks
16 or services;
17 (6) "Person", a natural person or any entity reco gnized as having legal
18 personhood under the laws of the state, explicitly excluding any AI system.
EXPLANA TION — Matter enclosed in bold-faced brackets [thus] in the above bill is not enacted and is
intended to be omitted from the law . Matter in bold-face type in the above bill is proposed language.
19 3. For all purposes under state law , AI systems ar e declar ed to be nonsentient
20 entities.
21 4. No AI system shall be granted the status of a person or any form of legal
22 personhood, nor be consider ed to possess consciousness, self-awar eness, or similar traits
23 of living beings.
24 5. No AI system shall be recogn ized as a spouse, domestic partner , or hold any
25 personal legal status analogous to marriage or union with a human or another AI
26 system. Any purported attempt to marry or cr eate a personal union with an AI system
27 is void and shall have no legal effect.
28 6. AI systems shall not be designated, appointed, or serve as any officer , director ,
29 manager , or similar ro le within any corporation, partnership, or other legal entity . Any
30 purported appointment of an AI system to such a role is void and has no legal effect.
31 7. AI systems shall not be recogn ized as legal entities capable of owning,
32 contr olling, or holding title to any form of prop erty including, but not limited to, r eal
33 estate, intellectual pro perty , financial accounts, and digital assets. All assets and
34 pr oprietary inter ests generated, managed, or otherwise associated with AI shall be
35 attributed to the human individuals or legally reco gnized organizations r esponsible for
36 the AI development, deployment, or operation.
37 8. Any dir ect or indirect harm caused by an AI system's operation, output, or
38 r ecommendation, when used as intended or misused, shall be the res ponsibility of the
39 deployer or user who directed or employed the AI. Developers or deployers may be held
40 liable if a defect in design, construction, or instructions for use of the AI system
41 pr oximately causes harm, consistent with prod uct liability principles; however , mer e
42 misuse or intentional wron gdoing by the user shall not impute liability to the developer
43 or the deployer absent pro of of negligence or design defects, nor shall wrong doing by the
44 developer impute liability to the developer . In cases of negligence, or gross negligence,
45 or malpractice, the developer or deployer shall be absolved of any criminal wron gdoing
46 or punitive damages, provi ded that the standards established in subsection 9 of this
47 section were followed.
48 9. Deployers and developers shall maintain pro per oversight and contr ol
49 measur es over any AI system whose outputs or r ecommendations could r easonably be
50 expected to impact human welfar e, prop erty , or public safety . Failur e to prov ide
51 adequate supervision or safeguards against for eseeable risks may constitute negligence
52 or another applicable basis of liability . For purposes of this subsection, pr oper oversight
53 is demonstrated when a developer or deployer adher es to Missouri standards for AI risk
54 management which shall be derived fr om and based on the National Institute of
55 Standards and T echnology or NIST AI Risk Management Framework NIST AI 100-1
HCS HBs 1746 & 1769 2
56 standards for AI and cybersecurity . If Missouri does not have curren t standards or is
57 silent on any aspect of AI risk management, a developer or deployer of AI shall have
58 demonstrated pr oper oversight thr ough adher ence to the latest version of the NIST AI
59 Risk Management Framework NIST AI 100-1, or its successor .
60 10. An AI system is not an entity capable of bearing fault or liability in its own
61 right. Any attempt to shift blame solely onto an AI system shall be void as liability
62 r emains with human actors or entities as described in this section.
63 1 1. Developers and deployers of AI systems shall prioritize safety mechanisms
64 designed to pr event or mitigate risk of dir ect harm to individuals or pr operty .
65 12. (1) Merel y labeling an AI system as "aligned", "ethically trained", or "value
66 locked" shall not excuse or diminish the deployer's or developer's liability for harm.
67 Deployers shall rem ain r esponsible for demonstrating adequate safety featur es and risk
68 contr ols commensurate with the AI's level of potential harm.
69 (2) Advances in software, automation, algorithmic decision-making, artificial
70 intelligence, and other digital technologies shall not be used as the basis for a deployer
71 or developer to engage in conduct that is proh ibited under the law .
72 13. Deployers or developers of AI systems involved in sever e incidents r esulting
73 in significant bodily harm, death, or major pro perty damage shall prompt ly notify the
74 r elevant authorities and comply with any subsequent investigations.
75 14. The prov isions of this section shall apply to all AI systems developed, owned,
76 deployed, or operated on or after August 28, 2026. A developer or deployer that has
77 formally adopted a policy to comply with this section and demonstrated an
7 8 implementation plan before August 28, 2026, shall not be subject to the r equir ements
79 of this section before Marc h 1, 2027.
✔
HCS HBs 1746 & 1769 3