Bill

Bill > HB2829


PA HB2829

PA HB2829
Providing for the establishment of standards, protections and transparency of artificial intelligence frameworks developed by frontier developers; imposing duties on the Pennsylvania Emergency Management Agency and the Attorney General; and imposing penalties.


summary

Introduced
10/05/2026
In Committee
10/05/2026
Crossed Over
Passed
Dead

Introduced Session

2025-2026 Regular Session

Bill Summary

Providing for the establishment of standards, protections and transparency of artificial intelligence frameworks developed by frontier developers; imposing duties on the Pennsylvania Emergency Management Agency and the Attorney General; and imposing penalties.

AI Summary

This bill, titled the Frontier Artificial Intelligence Safety, Transparency and Accountability Act, establishes comprehensive state-level safety, transparency, and accountability rules for advanced artificial intelligence systems starting January 1, 2028. It specifically targets "frontier developers"—creators of massive foundation models trained using unprecedented computing power (specifically exceeding 10^26 integer or floating-point operations)—and imposes stricter requirements on "large frontier developers," defined as those with annual gross revenues exceeding $500,000,000. Under the legislation, administered jointly by the Pennsylvania Emergency Management Agency (PEMA) and the Attorney General, large frontier developers must file annual disclosure statements, publish and implement an internal "frontier AI framework" detailing how they assess and mitigate catastrophic risks (such as assisting in the creation of weapons of mass destruction or causing mass casualties and extreme property damage), and undergo annual independent third-party audits. Furthermore, developers must publish transparency reports before deploying new or modified models, maintain strict cybersecurity to protect unreleased model weights, and report any "critical safety incidents" (such as a model evading human control or causing serious harm) to state authorities within 72 hours, or within 24 hours if there is an imminent risk of death or physical injury. To protect workers who report safety violations or dangerous risks, the bill includes robust whistleblower protections and internal anonymous reporting channels, while penalizing knowing misrepresentations, unpermitted deployment of dangerous models, and non-compliance with civil penalties of up to $1,000,000 for a first violation and $3,000,000 for subsequent offenses. Finally, the measure preempts local governments from enacting conflicting artificial intelligence regulations, ensures state-level oversight works in tandem with existing laws, and allows developers to satisfy state rules if they comply with substantially equivalent federal laws.

Committee Categories

Business and Industry

Sponsors (5)

Last Action

Communications & Technology (h) Hearing (10:00:00 10/19/2026 Room 205 Ryan Office) (on 10/19/2026)

Bill Topics

Banking, Finance, and Domestic Commerce
  • ‐ Consumer Safety and Consumer Fraud
Space, Science, Technology, and Communications
  • ‐ Internet and Computer Issues

bill text


bill summary

Loading...

bill summary

Loading...

bill summary

Document Type Source Location Created
State Bill Page https://www.palegis.us/legislation/bills/2025/hb2829 10/05/2026
BillText https://www.palegis.us/legislation/bills/text/PDF/2025/0/HB2829/PN4002 10/06/2026
Loading...