...II
119 THCONGRESS
1STSESSION S. 2081
To establish immunity from civil liability for certain artificial intelligence
developers, and for other purposes.
IN THE SENATE OF THE UNITED STATES
JUNE 12, 2025
Ms. L UMMIS introduced the following bill; which was read twice and referred
to the Committee on Commerce, Sc...
... the society 8
of the United States. 9
(3) The current lack of clarity regarding liabil- 10
ity for artificial intelligence errors creates uncer- 11
tainty that impedes the responsible integration of 12
these beneficial technologies into professional serv- 13
ices and economic activity. 14
(4) Many artificial intelligence...
...12
these beneficial technologies into professional serv- 13
ices and economic activity. 14
(4) Many artificial intelligence systems operate 15
with limited transparency regarding their capabili- 16
ties, limitations, and default instructions, making it 17
difficult for professional users to assess appropriate 18
use cases...
...frameworks, will promote responsible inno- 5
vation while ensuring that the benefits and risks of 6
artificial intelligence systems are properly under- 7
stood and managed as these technologies continue to 8
advance. 9
(7) The development of artificial intelligence 10
systems that may significantly impact the future of 11...
...ly under- 7
stood and managed as these technologies continue to 8
advance. 9
(7) The development of artificial intelligence 10
systems that may significantly impact the future of 11
human civilization warrants a governance approach 12
that balances innovation incentives with robust 13
transparency requirements and appropr...
...grams, trains, 10
modifies, or substantially contributes to the cre- 11
ation or modification of an artificial intelligence 12
product; 13
(B) exercises control over the design speci- 14
fications, functionality, capabilities, limitations, 15
or intended uses of an artificial intelligence 16
product; or 17
(C) markets, di...
...the design speci- 14
fications, functionality, capabilities, limitations, 15
or intended uses of an artificial intelligence 16
product; or 17
(C) markets, distributes, licenses, or makes 18
available an artificial intelligence product under 19
their own name, brand, or trademark, regard- 20
less of whether the person crea...
...ificial intelligence 16
product; or 17
(C) markets, distributes, licenses, or makes 18
available an artificial intelligence product under 19
their own name, brand, or trademark, regard- 20
less of whether the person creates the original 21
underlying technology of the artificial intel- 22
ligence product. 23
(4) E RROR .—...
...W3PROD with $$_JOB5
•S 2081 IS(A) any output, action, recommendation, 1
or material omission by an artificial intelligence 2
product that is false, misleading, fabricated, de- 3
ceptive, or incomplete in a manner that a rea- 4
sonable developer could foresee would cause 5
harm; or 6
(B) any failure of an artificial intel...
...manner that a rea- 4
sonable developer could foresee would cause 5
harm; or 6
(B) any failure of an artificial intelligence 7
product to perform a function or task that the 8
artificial intelligence product expressly or im- 9
plicitly represents itself as capable of per- 10
forming. 11
(5) L EARNED PROFESSIONAL .—The term...
... 6
(B) any failure of an artificial intelligence 7
product to perform a function or task that the 8
artificial intelligence product expressly or im- 9
plicitly represents itself as capable of per- 10
forming. 11
(5) L EARNED PROFESSIONAL .—The term 12
‘‘learned professional’’ means an individual who— 13
(A) possesses spec...
...itigations, including de- 7
tection, evaluation, management, and safeguards 8
against errors, of an artificial intelligence product. 9
(7) M ODEL SPECIFICATION .—The term ‘‘model 10
specification’’— 11
(A) means the text or other configuration 12
instructions of an artificial intelligence prod- 13
uct— 14
(i) supplied by ...
...term ‘‘model 10
specification’’— 11
(A) means the text or other configuration 12
instructions of an artificial intelligence prod- 13
uct— 14
(i) supplied by a developer; 15
(ii) that establish the intended base 16
behavior, tone, constraints, or goals of the 17
artificial intelligence product; and 18
(iii) that materially...
...eloper; 15
(ii) that establish the intended base 16
behavior, tone, constraints, or goals of the 17
artificial intelligence product; and 18
(iii) that materially influence the out- 19
puts of the artificial intelligence product 20
across users or sessions, including the sys- 21
tem prompt provided to the model before 22
e...
... 17
artificial intelligence product; and 18
(iii) that materially influence the out- 19
puts of the artificial intelligence product 20
across users or sessions, including the sys- 21
tem prompt provided to the model before 22
engaging with user queries; and 23
(B) includes— 24
VerDate Sep 11 2014 22:54 Jun 23, 2025 Jkt 05...
...; 4
(ii) any constitution or analogous 5
guiding document used when training or 6
fine-tuning of an artificial intelligence 7
product, including in automated schemes 8
in which an artificial intelligence system 9
trains another artificial intelligence system; 10
and 11
(iii) the instructions, rubrics, or other 12
guidance...
...6
fine-tuning of an artificial intelligence 7
product, including in automated schemes 8
in which an artificial intelligence system 9
trains another artificial intelligence system; 10
and 11
(iii) the instructions, rubrics, or other 12
guidance provided to human raters or eval- 13
uators of an artificial intelligence produ...
...oduct, including in automated schemes 8
in which an artificial intelligence system 9
trains another artificial intelligence system; 10
and 11
(iii) the instructions, rubrics, or other 12
guidance provided to human raters or eval- 13
uators of an artificial intelligence product 14
the feedback of whom is used to train or 1...
...) the instructions, rubrics, or other 12
guidance provided to human raters or eval- 13
uators of an artificial intelligence product 14
the feedback of whom is used to train or 15
fine-tune the artificial intelligence product. 16
SEC. 4. CONDITIONAL IMMUNITY FROM CIVIL LIABILITY 17
FOR ARTIFICIAL INTELLIGENCE DEVEL- 18
OPE...
... of an artificial intelligence product 14
the feedback of whom is used to train or 15
fine-tune the artificial intelligence product. 16
SEC. 4. CONDITIONAL IMMUNITY FROM CIVIL LIABILITY 17
FOR ARTIFICIAL INTELLIGENCE DEVEL- 18
OPERS. 19
(a) S AFE HARBOR ELIGIBILITY .—A developer shall 20
be immune from civil liability for...
...BOR ELIGIBILITY .—A developer shall 20
be immune from civil liability for errors generated by an 21
artificial intelligence product when used by a learned pro-22
fessional in the course of providing professional services 23
to a client if the developer— 24
VerDate Sep 11 2014 22:54 Jun 23, 2025 Jkt 059200 PO 00000 Frm 000...
...i) only relating to information that 9
would reveal trade secrets unrelated to the 10
safety of the artificial intelligence product; 11
and 12
(ii) only if the developer furnishes 13
contemporaneously with each redaction a 14
written justification for the redaction iden- 15
tifying the basis for withholding the infor- 16
...
...onals describing the known 19
limitations, failure modes, and appropriate domains 20
of use for the artificial intelligence product. 21
(b) S COPE OF IMMUNITY .—The immunity provided 22
under subsection (a) shall be conferred to a developer only 23
for acts or omissions that do not constitute recklessness 24
or willful mi...
...BW3PROD with $$_JOB9
•S 2081 IS(c) D UTY TOUPDATE .—Immunity under subsection 1
(a) relating to an artificial intelligence product shall not 2
apply to a developer— 3
(1) that does not update the model card, model 4
specification, and documentation with respect to the 5
artificial intelligence product as described in sub...
...that does not update the model card, model 4
specification, and documentation with respect to the 5
artificial intelligence product as described in sub- 6
section (a)(1) by the date that is 30 days after the 7
date on which the developer— 8
(A) deploys a new version of the artificial 9
intelligence product; or 10
(B) disc...
...ficial 9
intelligence product; or 10
(B) discovers a new and material failure 11
mode affecting the artificial intelligence prod- 12
uct; and 13
(2) of which the failure to make an update de- 14
scribed in paragraph (1) by the applicable date de- 15
scribed in that paragraph proximately causes a harm 16
occurring after th...
...any claim arising under State law against 20
a developer for an error arising from the use of an 21
artificial intelligence product by a learned profes- 22
sional in providing professional services if the devel- 23
oper is immune from civil liability under subsection 24
(a). 25
VerDate Sep 11 2014 22:54 Jun 23, 2025 Jkt 0...
...sed on fraud, knowing 3
misrepresentation, or conduct outside the scope of 4
professional use of an artificial intelligence product 5
by a learned professional. 6
SEC. 5. PRESERVATION OF OTHER IMMUNITIES AND PRIVI-7
LEGES. 8
Nothing in this Act shall be construed to affect any 9
immunity from civil liability established b...
... document used when training or 6
fine-tuning of an artificial intelligence 7
product, including in automated schemes 8
in which an artificial intelligence system 9
trains another artificial intelligence system; 10
and 11
(iii) the instructions, rubrics, or other 12
guidance provided to human raters or eval-...
5022 results:
Description: A bill to establish immunity from civil liability for certain artificial intelligence developers, and for other purposes.
Summary: The RISE Act of 2025 establishes civil liability immunity for AI developers when their products are used by professionals, contingent on transparency and accountability measures, to encourage responsible AI innovation.
Collection: Legislation
Status date: June 12, 2025
Status: Introduced
Primary sponsor: Cynthia Lummis
(sole sponsor)
Last action: Read twice and referred to the Committee on Commerce, Science, and Transportation. (June 12, 2025)
Keywords (occurrence):
artificial intelligence (31)
automated (1)
show keywords in context
Summary: Senate Amendment 1045 proposes a deficit-neutral reserve fund to advance border technologies, including AI for detecting smuggling and assisting vulnerable populations, ensuring no deficit increase from 2025-2034.
Collection: Congressional Record
Status date: Feb. 20, 2025
Status: Issued
Source: Congress
Keywords (occurrence):
automated (2)
show keywords in context
Description: Crimes: other; use of artificial intelligence to commit certain crimes; prohibit. Amends 1931 PA 328 (MCL 750.1 - 750.568) by adding sec. 411y.
Summary: The bill criminalizes the use of artificial intelligence (AI) to facilitate crimes in Michigan, imposing felony charges and mandatory prison sentences for offenders while ensuring consecutive sentences for related crimes.
Collection: Legislation
Status date: June 24, 2025
Status: Introduced
Primary sponsor: Sarah Lightner
(sole sponsor)
Last action: Bill Electronically Reproduced 06/24/2025 (June 25, 2025)
Keywords (occurrence):
automated (2)
show keywords in context
Description: Joint Order, to Carry Over Certain Matters from the First Special Session of the 132nd Legislature
Summary: The bill carries over various legislative matters from the First Special Session of the 132nd Legislature for consideration in future sessions, addressing topics such as transportation, healthcare, taxation, and housing.
Collection: Legislation
Status date: June 25, 2025
Status: Introduced
Primary sponsor: Teresa Pierce
(sole sponsor)
Last action: In concurrence. ORDERED SENT FORTHWITH. (June 25, 2025)
Keywords (occurrence):
artificial intelligence (1)
automated (1)
Description: Trade: business regulation; requirements and safety standards for developers of certain artificial intelligence models; provide for. Creates new act.
Summary: The bill establishes safety and security protocols for large developers of artificial intelligence models, aiming to manage critical risks, protect employees, and enforce compliance through civil sanctions and auditing.
Collection: Legislation
Status date: June 24, 2025
Status: Introduced
Primary sponsor: Sarah Lightner
(sole sponsor)
Last action: Bill Electronically Reproduced 06/24/2025 (June 25, 2025)
Keywords (occurrence):
artificial intelligence (2)
foundation model (14)
show keywords in context
Description: Crimes: other; use of artificial intelligence to commit certain crimes; prohibit.
Summary: House Bill No. 4667 prohibits using artificial intelligence to commit crimes in Michigan, establishing felony charges and mandatory prison sentences for those who develop or utilize AI for illegal purposes.
Collection: Legislation
Status date: June 24, 2025
Status: Introduced
Primary sponsor: Sarah Lightner
(sole sponsor)
Last action: Bill Electronically Reproduced 06/24/2025 (June 25, 2025)
Keywords (occurrence):
automated (2)
show keywords in context
Description: Trade: business regulation; requirements and safety standards for developers of certain artificial intelligence models; provide for.
Summary: The bill establishes safety and security protocols for large developers of artificial intelligence models to manage critical risks, promote employee protections, and enforce compliance through civil sanctions.
Collection: Legislation
Status date: June 24, 2025
Status: Introduced
Primary sponsor: Sarah Lightner
(sole sponsor)
Last action: Bill Electronically Reproduced 06/24/2025 (June 25, 2025)
Keywords (occurrence):
artificial intelligence (2)
foundation model (14)
show keywords in context
Summary: The bill proposes amendments to the congressional budget for fiscal years 2025-2034, establishing deficit-neutral reserve funds to enhance healthcare access, support rural hospitals, and promote various community and infrastructure improvements.
Collection: Congressional Record
Status date: Feb. 20, 2025
Status: Issued
Source: Congress
Keywords (occurrence):
artificial intelligence (2)
automated (2)
show keywords in context
Summary: Senate Amendment 1069 establishes a deficit-neutral reserve fund to adjust budget allocations for enforcing export controls against Chinese entities, specifically targeting access to U.S. artificial intelligence chips.
Collection: Congressional Record
Status date: Feb. 20, 2025
Status: Issued
Source: Congress
Keywords (occurrence):
artificial intelligence (1)
show keywords in context
Summary: The "Protecting Patients from Deceptive Drug Ads Act" aims to regulate and penalize false or misleading communications about prescription drugs by social media influencers and healthcare providers to protect patients.
Collection: Congressional Record
Status date: Feb. 20, 2025
Status: Issued
Source: Congress
Keywords (occurrence):
artificial intelligence (1)
show keywords in context
Summary: The "Protecting Patients from Deceptive Drug Ads Act" regulates communications about prescription drugs, imposing penalties for false advertising by influencers and healthcare providers, aiming to ensure accurate information is conveyed.
Collection: Congressional Record
Status date: Feb. 20, 2025
Status: Issued
Source: Congress
Keywords (occurrence):
artificial intelligence (1)
show keywords in context
Description: An act to add Chapter 40 (commencing with Section 22949.85) to Division 8 of the Business and Professions Code, relating to business.
Summary: The Preventing Algorithmic Price Fixing Act prohibits the sale or use of price-fixing algorithms for setting prices or rental levels, allowing civil penalties for violations.
Collection: Legislation
Status date: Feb. 14, 2025
Status: Introduced
Primary sponsor: Aisha Wahab
(sole sponsor)
Last action: Read second time and amended. Re-referred to Com. on APPR. (April 24, 2025)
Keywords (occurrence):
artificial intelligence (3)
algorithm (11)
show keywords in context
Description: An act to add Section 1947.16 to the Civil Code, relating to housing.
Summary: Senate Bill No. 52 prohibits the use and distribution of rental pricing algorithms that manipulate housing rates, aiming to protect renters from inflated prices and ensure fair competition in California's rental market.
Collection: Legislation
Status date: Dec. 20, 2024
Status: Introduced
Primary sponsor: Sasha Renee Perez
(sole sponsor)
Last action: Read second time and amended. Re-referred to Com. on APPR. (April 24, 2025)
Category:
Societal Impact
Data Governance
System Integrity (see reasoning)
Sector:
Government Agencies and Public Services
Private Enterprises, Labor, and Employment (see reasoning)
Keywords (occurrence): artificial intelligence (1) algorithm (16) show keywords in context
Societal Impact
Data Governance
System Integrity (see reasoning)
This text primarily discusses issues surrounding the use of rental pricing algorithms in the housing market, detailing their impact on rental rates and the overall housing market. As such, it aligns closely with several categories. For Social Impact, the legislation tackles societal concerns such as inflated rental pricing and market manipulation, affecting a significant demographic of renters. For Data Governance, the text emphasizes the importance of appropriate data use and mandates against using nonpublic competitor data, which relates directly to the management of data in AI systems. The System Integrity category is relevant, as it discusses unlawful practices that ensure fair competition and prevents misuse of algorithms. However, the aspects of Robustness are less relevant, as the legislation focuses more on prohibitions rather than benchmarks for AI performance or audits. Overall, the text shows significant relevance to Social Impact and Data Governance, moderate relevance to System Integrity, and low relevance to Robustness.
Sector:
Government Agencies and Public Services
Private Enterprises, Labor, and Employment (see reasoning)
The text discusses housing and rental rates, with a clear focus on the ramifications of algorithmic pricing on the rental market. It's particularly relevant to the Private Enterprises, Labor, and Employment sector as it investigates how corporate practices utilize AI technologies in housing, which directly impact landlords and renters. Additionally, it contains implications for Government Agencies and Public Services, particularly regarding enforcement mechanisms introduced by the Attorney General and local authorities. Therefore, it would score high in these sectors. The elements concerning AI do not directly involve Healthcare, Politics and Elections, Judicial System, Academic and Research Institutions, International Cooperation and Standards, Nonprofits and NGOs, or Hybrid, Emerging, and Unclassified sectors. Hence, this text scores high relevance in the Private Enterprises sector and moderate relevance in the Government Agencies sector.
Keywords (occurrence): artificial intelligence (1) algorithm (16) show keywords in context
Description: Memorializes the United States Congress to take action relative to the use of artificial intelligence in health care
Summary: The bill urges Congress to regulate artificial intelligence in healthcare, advocating for physician control over medical decisions and ensuring patient safety, data privacy, and ethical standards.
Collection: Legislation
Status date: June 10, 2025
Status: Passed
Primary sponsor: Jeremy LaCombe
(3 total sponsors)
Last action: Taken by the Clerk of the House and presented to the Secretary of State in accordance with the Rules of the House. (June 11, 2025)
Keywords (occurrence):
machine learning (1)
show keywords in context
Description: Prescribes requirements and safeguards for the use of an artificial intelligence, algorithm, or other software tool for the purpose of utilization review for health and accident insurance.
Summary: This bill mandates requirements and safeguards for using artificial intelligence in health insurance utilization reviews, ensuring accurate, fair assessments while maintaining human oversight and patient data protection.
Collection: Legislation
Status date: May 13, 2025
Status: Introduced
Primary sponsor: Kristen Gonzalez
(sole sponsor)
Last action: REFERRED TO HEALTH (May 13, 2025)
Keywords (occurrence):
artificial intelligence (22)
algorithm (31)
show keywords in context
Description:
Summary:
Collection:
Status date:
Status:
Primary sponsor:
( total sponsors)
Source:
Last action: ()
Category:
None
(see reasoning)
Sector: None (see reasoning)
Keywords (occurrence): artificial intelligence () machine learning () neural network () deep learning () automated () deepfake () synthetic media () large language model () foundation model () chatbot () recommendation system () algorithm () autonomous vehicle ()
Sector: None (see reasoning)
Keywords (occurrence): artificial intelligence () machine learning () neural network () deep learning () automated () deepfake () synthetic media () large language model () foundation model () chatbot () recommendation system () algorithm () autonomous vehicle ()
Description: Relative to autonomous vehicles. Transportation.
Summary: This bill outlines regulations for operating fully autonomous vehicles in Massachusetts, including licensing, safety standards, and insurance requirements, aiming to facilitate their integration into the transportation system.
Collection: Legislation
Status date: Feb. 27, 2025
Status: Introduced
Primary sponsor: Daniel Cahill
(2 total sponsors)
Last action: Senate concurred (Feb. 27, 2025)
Keywords (occurrence):
automated (25)
autonomous vehicle (25)
show keywords in context
Description:
Summary:
Collection:
Status date:
Status:
Primary sponsor:
( total sponsors)
Source:
Last action: ()
Category:
None
(see reasoning)
Sector: None (see reasoning)
Keywords (occurrence): artificial intelligence () machine learning () neural network () deep learning () automated () deepfake () synthetic media () large language model () foundation model () chatbot () recommendation system () algorithm () autonomous vehicle ()
Sector: None (see reasoning)
Keywords (occurrence): artificial intelligence () machine learning () neural network () deep learning () automated () deepfake () synthetic media () large language model () foundation model () chatbot () recommendation system () algorithm () autonomous vehicle ()
Description: Requires certain political communications to include provenance data for all audios, images or videos used in such communications; provides for the repeal of certain provisions upon the expiration thereof.
Summary: The bill mandates that political communications include provenance data for audio, images, or videos, detailing their origin and any modifications, aiming to enhance transparency in media usage during elections.
Collection: Legislation
Status date: May 15, 2025
Status: Introduced
Primary sponsor: Kristen Gonzalez
(sole sponsor)
Last action: REFERRED TO ELECTIONS (May 15, 2025)
Keywords (occurrence):
artificial intelligence (5)
show keywords in context
Description:
Summary:
Collection:
Status date:
Status:
Primary sponsor:
( total sponsors)
Source:
Last action: ()
Category:
None
(see reasoning)
Sector: None (see reasoning)
Keywords (occurrence): artificial intelligence () machine learning () neural network () deep learning () automated () deepfake () synthetic media () large language model () foundation model () chatbot () recommendation system () algorithm () autonomous vehicle ()
Sector: None (see reasoning)
Keywords (occurrence): artificial intelligence () machine learning () neural network () deep learning () automated () deepfake () synthetic media () large language model () foundation model () chatbot () recommendation system () algorithm () autonomous vehicle ()