Bill Summary – FastDemocracy

BILL AS INTRODUCED                                                        H.711
                                                              Page 1 of 14
                                      H.711
 Introduced by Representatives Priestley of Bradford, Burrows of West
                Windsor, Chase of Chester, Christie of Hartford, Jerome of
                Brandon, Masland of Thetford, Roberts of Halifax, Sibilia of
                Dover, Sims of Craftsbury, Templeman of Brownington, White
                of Bethel, and Williams of Barre City
 Referred to Committee on
 Date:
 Subject: Commerce and trade; consumer protection; liability for developers
          and deployers of artificial intelligence systems
 Statement of purpose of bill as introduced: This bill proposes to create
 oversight and liability standards for developers and deployers of inherently
 dangerous artificial intelligence systems.
    An act relating to creating oversight and liability standards for developers
    and deployers of inherently dangerous artificial intelligence systems
 It is hereby enacted by the General Assembly of the State of Vermont:
 Sec. 1. 9 V.S.A. chapter 63, subchapter 12 is added to read:
          Subchapter 12. Artificial Intelligence Oversight and Liability
 § 2495a. LEGISLATIVE INTENT
                                                                       VT LEG #373232 v.1
     BILL AS INTRODUCED                                                            H.711
                                                                  Page 2 of 14
    (a) Artificial intelligence systems are products that shift decision-making
 power and responsibility away from persons to software-based systems, often
 without direct human oversight. An artificial intelligence system can be
 inherently dangerous due to its capabilities, potential for misuse or
 exploitation, and ability to unilaterally evolve.
    (b) Developers of sophisticated artificial intelligence systems have an
 obligation to make such systems safe when used in reasonably foreseeable
 ways. Deployers of these products also have an obligation to ensure that the
 products are safe and used in a way that does not materially affect an
 individual’s rights.
    (c) In the artificial intelligence ecosystem, there will typically be multiple
 suppliers upstream of a consumer. The original developer of an artificial
 intelligence system should be responsible for harms attributable to the artificial
 intelligence system, even if the developer is not the deployer of the system to a
 consumer. Small businesses using off-the-shelf artificial intelligence products
 according to the product’s terms of use are not intended to be covered by this
 act.
 § 2495b. DEFINITIONS
    As used in this subchapter:
        (1) “Artificial intelligence system” means a machine-based system that
 can, for a given set of objectives, make predictions, recommendations, or VT LEG #373232 v.1
     BILL AS INTRODUCED                                                            H.711
                                                                  Page 3 of 14
 decisions influencing real or virtual environments. Artificial intelligence
 systems use machine- and human-based inputs to perceive real and virtual
 environments, abstract such perceptions into models through analysis in an
 automated manner, and use model inference to formulate options for
 information or action.
       (2) “Biometric data” means data that depict or describe physical,
 biological, or behavioral traits, characteristics, or measurements of or relating
 to an identified or identifiable person’s body. Biometric information includes
 depictions, images, descriptions, or recordings of an individual’s facial
 features, iris or retina, finger or handprints, voice, genetics, or characteristic
 movements or gestures. Biometric information also includes data derived from
 such depictions, images, descriptions, or recordings, to the extent that it would
 be reasonably possible to identify the person from whose information the data
 had been derived.
       (3) “Consequential decision” means a decision that either has a legal or
 similarly significant effect on an individual’s access to the criminal justice
 system, housing, employment, credit, education, health care, or insurance.
       (4) “Consumer” means any individual who is a resident of this State.
       (5) “Deployer” means a person, including a developer, who uses or
 operates an artificial intelligence system for internal use or for use by third
 parties in the State.
                                                                           VT LEG #373232 v.1
     BILL AS INTRODUCED                                                          H.711
                                                                Page 4 of 14
       (6) “Developer” means a person who designs, codes, produces, owns, or
 substantially modifies an artificial intelligence system for internal use or for
 use by a third party in the State.
       (7) “Dual-use foundational model” means an artificial intelligence
 system that:
          (A) is trained on broad data;
          (B) generally uses self-supervision;
          (C) contains at least 10 billion parameters;
          (D) is applicable across a wide range of contexts; and
          (E) exhibits, or could be easily modified to exhibit, high levels of
 performance at tasks that pose a serious risk to economic security, public
 health or safety, or any combination of those matters, such as by:
             (i) substantially lowering the barrier of entry for nonexperts to
 design, synthesize, acquire, or use chemical, biological, radiological, or nuclear
 (CBRN) weapons;
             (ii) enabling powerful offensive cyber operations through
 automated vulnerability discovery and exploitation against a wide range of
 potential targets of cyber attacks; or
             (iii) permitting the evasion of human control or oversight through
 means of deception or obfuscation.
                                                                         VT LEG #373232 v.1
     BILL AS INTRODUCED                                                        H.711
                                                              Page 5 of 14
       (8) “Generative artificial intelligence system” means an artificial
 intelligence system that is capable of generating output including text,
 imagery, audio, and synthetic data.
       (9) “High-risk artificial intelligence system” means any artificial
 intelligence system, regardless of the number of parameters and supervision
 structure, that is:
           (A) used, reasonably foreseeable as being used, or is a controlling
 factor in making a consequential decision;
           (B) used, or reasonably foreseeable as being used, to categorize
 groups of persons by sensitive and protected characteristics, such as race,
 ethnic origin, or religious belief;
           (C) used, or reasonably foreseeable as being used, in the direct
 management or operation of critical infrastructure;
           (D) used, or reasonably foreseeable as being used, in vehicles,
 medical devices, or in the safety system of a product;
           (E) used, or reasonably foreseeable as being used, to influence
 elections or voters; or
           (F) used to collect the biometric data of an individual from a
 biometric identification system without consent.
                                                                       VT LEG #373232 v.1
     BILL AS INTRODUCED                                                         H.711
                                                               Page 6 of 14
          (10) “Inherently dangerous artificial intelligence system” means a high-
 risk artificial intelligence system, dual-use foundational model, or generative
 artificial intelligence system.
 § 2495c. OVERSIGHT AND ENFORCEMENT
    (a) The Division of Artificial Intelligence within the Agency of Digital
 Services shall collect and review Artificial Intelligence Safety and Impact
 Assessments pursuant to this subchapter.
    (b) The Attorney General shall enforce the provisions of this subchapter
 and may bring an action in the name of the State against a deployer or
 developer for noncompliance to restrain by temporary or permanent injunction
 the noncompliance. The action may be brought in the Superior Court of the
 county in which such person resides, has a place of business, or is doing
 business. Said courts are authorized to issue temporary or permanent
 injunctions to restrain and prevent violations of this subchapter, such
 injunctions to be issued without bonds, or to dissolve, or revoke the certificate
 of authority of, a deployer or developer.
 § 2495d. ARTIFICIAL INTELLIGENCE SYSTEM SAFETY AND IMPACT
             ASSESSMENT
    (a) Each deployer of an inherently dangerous artificial intelligence system
 shall:
                                                                        VT LEG #373232 v.1
     BILL AS INTRODUCED                                                          H.711
                                                                Page 7 of 14
       (1) submit to the Division of Artificial Intelligence an Artificial
 Intelligence System Safety and Impact Assessment prior to deploying the
 inherently dangerous artificial intelligence system in this State, and every two
 years thereafter; and
       (2) submit to the Division of Artificial Intelligence an updated Artificial
 Intelligence System Safety and Impact Assessment if the deployer makes a
 material and substantial change to the inherently dangerous artificial
 intelligence system that includes:
          (A) the purpose for which the system is used for; or
          (B) the type of data the system processes or uses for training
 purposes.
    (b) Each Artificial Intelligence System Impact Assessment pursuant to
 subsection (a) of this section shall include, with respect to the inherently
 dangerous artificial intelligence system:
       (1) the purpose of the system;
       (2) the deployment context and intended use cases;
       (3) the benefits of use;
       (4) any foreseeable risk of unintended or unauthorized uses and the steps
 taken, to the extent reasonable, to mitigate such risk;
       (5) whether the model is proprietary;
                                                                         VT LEG #373232 v.1
     BILL AS INTRODUCED                                                          H.711
                                                                Page 8 of 14
       (6) a description of the data the system processes or uses for training
 purposes;
       (7) a description of transparency measures, including identifying to
 individuals when the system is in use;
       (8) identification of any third-party artificial intelligence systems or
 datasets the deployer relies on to train or operate the system, if applicable;
       (9) whether the developer of the system, if different than the deployer,
 disclosed the information pursuant to subsection 2495e(b) of this chapter as
 well as the results of testing, vulnerabilities, and the parameters for safe and
 intended use;
       (10) a description of the data the system, once deployed, processes as
 inputs;
       (11) a description of postdeployment monitoring and user safeguards,
 including a description of the oversight process in place to address issues as
 issues arise; and
       (12) a description of how the model impacts consequential decisions or
 the collection of biometric data.
    (c) Each deployer of a high-risk artificial intelligence system must submit a
 one-, six-, and 12-month testing result to the Division of Artificial Intelligence
 showing the reliability of the results generated by the system, any variance in
                                                                         VT LEG #373232 v.1
     BILL AS INTRODUCED                                                          H.711
                                                                Page 9 of 14
 those results over the testing periods, and any mitigation strategies for
 variances, in the first year of deployment.
    (d) Upon the Division of Artificial Intelligence receiving notice that a
 deployer of an inherently dangerous artificial intelligence system is not in
 compliance with the requirements under this section, the Division shall
 immediately notify the deployer of the finding in writing and order the
 deployer to submit the assessment required pursuant to subsection (a) of this
 section. If the deployer fails to submit the assessment within 45 days after
 receiving the notice, the Division of Artificial Intelligence shall notify the
 Attorney General in writing of the violation.
 § 2495e. STANDARD OF CARE
    (a) Each developer or deployer of any inherently dangerous artificial
 intelligence system that could be reasonably expected to impact consumers
 shall exercise reasonable care to avoid any reasonably foreseeable risk arising
 out of the development, intentional and substantial modification, or
 deployment of an artificial intelligence system that causes or is likely to cause:
       (1) the commission of a crime or unlawful act;
       (2) any unfair or deceptive treatment of or unlawful impact on an
 individual;
       (3) any physical, financial, relational, or reputational injury on an
 individual;
                                                                         VT LEG #373232 v.1
     BILL AS INTRODUCED                                                           H.711
                                                                Page 10 of 14
          (4) psychological injuries that would be highly offensive to a reasonable
 person;
          (5) any physical or other intrusion upon the solitude or seclusion, or the
 private affairs or concerns, of a person if such intrusion would be offensive to a
 reasonable person;
          (6) any violation to the intellectual property rights of persons under
 applicable State and federal laws;
          (7) discrimination on the basis of a person’s or class of person’s actual
 or perceived race, color, ethnicity, sex, sexual orientation, gender identity, sex
 characteristics, religion, national origin, familial status, biometric information,
 or disability status;
          (8) distortion of a person’s behavior in a manner that causes or is likely
 to cause that person or another person physical or psychological harm; or
          (9) the exploitation of the vulnerabilities of a specific group of persons
 due to their age or physical or mental disability in order to materially distort
 the behavior of a person pertaining to that group in a manner that causes or is
 likely to cause that person or another person physical or psychological harm.
    (b) Each developer of an inherently dangerous artificial intelligence system
 shall:
          (1) document and disclose to any actual or potential deployer of the
 artificial intelligence system any reasonably foreseeable risk, including by
                                                                           VT LEG #373232 v.1
     BILL AS INTRODUCED                                                           H.711
                                                                Page 11 of 14
 unintended or unauthorized uses, that causes or is likely to cause any of the
 injuries as set forth in subsection (a) of this section; and
       (2) document and disclose to any actual or potential deployer of the
 artificial intelligence system any risk mitigation processes that are reasonably
 foreseeable to mitigate any injury as set forth in subsection (a) of this section.
 § 2495f. UNSAFE ARTIFICIAL INTELLIGENCE PRODUCTS,
           PROHIBITIONS
    (a) No developer shall offer, sell, lease, give, or otherwise place in the
 stream of commerce:
       (1) an inherently dangerous artificial intelligence system, unless the
 developer has conducted a documented testing, evaluation, verification, and
 validation of that system at least as stringent as the latest version of the
 Artificial Intelligence Risk Management Framework published by the National
 Institute of Standards and Technology (NIST); or
       (2) an artificial intelligence system that creates reasonably foreseeable
 risks pursuant to section 2495e of this chapter, unless the developer mitigates
 these risks to the extent possible, considers alternatives, and discloses
 vulnerabilities and mitigation tactics to a deployer.
    (b) No deployer shall deploy an inherently dangerous artificial intelligence
 system or an artificial intelligence system that creates reasonably foreseeable
 risks pursuant to section 2495e of this chapter unless the deployer has designed VT LEG #373232 v.1
     BILL AS INTRODUCED                                                           H.711
                                                                Page 12 of 14
 and implemented a risk management policy and program for such model or
 system. The risk management policy shall specify the principles, processes,
 and personnel that the deployer shall use in maintaining the risk management
 program to identify, mitigate, and document any risk that is a reasonably
 foreseeable consequence of deploying or