HB 1533 — An Act amending Title 18 (Crimes and Offenses) of the Pennsylvania Consolidated Statutes, in culpability, providing for liability for deployment of artificial intelligence system.

PRINTER'S NO.     1794

                     THE GENERAL ASSEMBLY OF PENNSYLVANIA



                          HOUSE BILL
                          No. 1533
                                                Session of
                                                  2025

     INTRODUCED BY HOWARD, PIELLI, SANCHEZ, HILL-EVANS, VENKAT,
        WEBSTER AND NEILSON, MAY 30, 2025

     REFERRED TO COMMITTEE ON JUDICIARY, MAY 30, 2025


                                     AN ACT
 1   Amending Title 18 (Crimes and Offenses) of the Pennsylvania
 2      Consolidated Statutes, in culpability, providing for
 3      liability for deployment of artificial intelligence system.
 4      The General Assembly of the Commonwealth of Pennsylvania
 5   hereby enacts as follows:
 6      Section 1.     Title 18 of the Pennsylvania Consolidated
 7   Statutes is amended by adding a section to read:
 8   § 306.1.    Liability for deployment of artificial intelligence
 9                  system.
10      (a)     Liability.--A person that engages in the deployment of
11   an artificial intelligence system to perform an automated or
12   autonomous function shall be subject to criminal or civil
13   liability, or both, for any negative outcome arising from the
14   deployment, including:
15            (1)   Physical harm or endangerment caused by artificial
16      intelligence system-driven machinery, robotics, self-driving
17      vehicles or autonomous systems.
18            (2)   Economic misconduct, including price-fixing, bid-
 1      rigging, market allocation or fraud conducted through
 2      artificial intelligence system-driven decision making.
 3            (3)   Unlawful data scraping, copyright infringement,
 4      intellectual property theft or violations of privacy and data
 5      protection laws committed through automated or autonomous
 6      systems.
 7            (4)   Discriminatory, defamatory or otherwise harmful
 8      decision making arising from algorithmic bias or automated
 9      processing.
10            (5)   False, deceptive or misleading statements generated
11      or distributed through artificial intelligence system models
12      under the control of a person.
13            (6)   Any other act that, if committed by a person, would
14      constitute a violation of law or amount to tortious conduct.
15      (b)   Automation.--Automating an action does not remove a
16   person's general duty of care relating to the action as
17   established under common law.
18      (c)   Defenses.--
19            (1)   Except as provided in paragraph (2), a person may
20      not avoid liability under this section by asserting that:
21                  (i)    The artificial intelligence system acted
22            autonomously or without explicit human instruction after
23            the person engaged in the deployment of the artificial
24            intelligence system.
25                  (ii)    The outcomes of the artificial intelligence
26            system's decision making were unintended, unforeseen or
27            the result of machine-learning adaptation.
28                  (iii)    The artificial intelligence system was
29            trained, developed or provided by a third party and the
30            person merely initiated the deployment of the artificial

20250HB1533PN1794                       - 2 -
 1            intelligence system.
 2                  (iv)    The artificial intelligence system was
 3            marketed, certified or believed to be "safe," "self-
 4            regulating" or "autonomous."
 5            (2)   A person may assert a defense to liability under
 6      paragraph (1) by demonstrating that:
 7                  (i)    the person implemented reasonable, ongoing
 8            oversight, safeguards and fail-safe mechanisms designed
 9            to prevent unlawful, negligent or harmful conduct and
10            took timely corrective action once the person became
11            aware of a risk or failure relating to the artificial
12            intelligence system; or
13                  (ii)    the artificial intelligence system's harmful
14            act or omission resulted solely from an unforeseeable and
15            unauthorized interference by an external actor and the
16            person had previously implemented reasonable security
17            measures designed to prevent the unauthorized use.
18            (3)   The failure to establish a defense under paragraph
19      (2) shall result in full liability for the deployment as
20      specified in this section.
21      (d)   Definitions.--As used in this section, the following
22   words and phrases shall have the meanings given to them in this
23   subsection unless the context clearly indicates otherwise:
24      "Artificial intelligence system."       As follows:
25            (1)   A machine-based system that can, for a given set of
26      human-defined objectives, make predictions, recommendations
27      or decisions influencing real or virtual environments,
28      including the ability to:
29                  (i)    Perceive real and virtual environments.
30                  (ii)    Abstract perceptions made under this paragraph

20250HB1533PN1794                       - 3 -
 1        into models through analysis in an automated manner.
 2              (iii)   Use model inference to formulate options for
 3        information or action based on outcomes under
 4        subparagraphs (i) and (ii).
 5        (2)   The term includes generative artificial intelligence
 6    and any substantially similar technology yet to be developed.
 7    "Deployment."     As follows:
 8        (1)   The use of an artificial intelligence system in a
 9    manner that has the potential to affect external persons,
10    systems or legal interests.
11        (2)   The term includes commercial implementation,
12    enterprise use, individual use or the use of autonomous
13    systems affecting third parties.
14        (3)   The term does not include the use of purely
15    experimental, nondeployed research models if the models are
16    reasonably secured to prevent unauthorized access and
17    deployment.
18    Section 2.    This act shall take effect in 60 days.




20250HB1533PN1794                     - 4 -