Laws & Statutes › MN › MN-HF-5051
HF-5051
MN · State · USA
MN
USA
● Pending
Proposed Effective Date
2026-08-01
Minnesota H.F. No. 5051 — A bill for an act relating to consumer protection; requiring disclosures when selling or distributing programs with artificial intelligence; proposing coding for new law in Minnesota Statutes, chapter 325G
This short bill would require any seller or distributor of a program containing artificial intelligence technology to make five categories of pre-sale disclosures: the AI's manufacturers or creators, technical support contact information, the AI's functions, the types of modeling used, and all safety features including human-in-the-loop integration. The bill uses a broad definition of AI aligned with the NIST framework. It does not specify an enforcement mechanism, penalties, or remedies, and does not define 'seller or distributor' or 'program' as formal terms. The bill was referred to the Committee on Commerce Finance and Policy on April 22, 2026.
Passage Likelihood
Low
Chamber No passage
Committee No action
Majority party No
Bipartisan No
Prior session None
Legislative History
2026-04-22 introduced Introduction and first reading, referred to Commerce Finance and Policy
2026-04-23 other Author added Hollins
Mapped Requirements
G-02 Public Transparency & Documentation
Official Sources
Full Text ↗
Entry Last Reviewed
2026-05-10
AI generated
Summary

This short bill would require any seller or distributor of a program containing artificial intelligence technology to make five categories of pre-sale disclosures: the AI's manufacturers or creators, technical support contact information, the AI's functions, the types of modeling used, and all safety features including human-in-the-loop integration. The bill uses a broad definition of AI aligned with the NIST framework. It does not specify an enforcement mechanism, penalties, or remedies, and does not define 'seller or distributor' or 'program' as formal terms. The bill was referred to the Committee on Commerce Finance and Policy on April 22, 2026.

Enforcement & Penalties
Enforcement Authority
No enforcement authority is specified in the bill. The bill proposes new law under Minnesota Statutes chapter 325G (consumer protection) but does not designate an enforcing agency, create a private right of action, or specify any enforcement mechanism. Violations would potentially fall under the Minnesota Attorney General's general consumer protection enforcement authority under existing chapter 325G provisions, but the bill itself is silent on enforcement.
Penalties
The bill does not specify any penalties, damages, or remedies for violations. No civil penalties, statutory damages, injunctive relief provisions, or attorney fee provisions are included.
Who Is Covered
Compliance Obligations 1 obligation · click obligation ID to open requirement page
G-02 Public Transparency & Documentation · G-02.1 · DeveloperDeployerDistributor · General Consumer App
Minn. Stat. § 325G.64, Subd. 2
Plain Language
Before selling or distributing any program that contains AI, the seller or distributor must disclose five categories of information to the buyer or recipient: (1) the business names of the AI's manufacturers or creators, (2) contact information for technical support, (3) the functions the AI performs, (4) the types of modeling the AI uses, and (5) all safety features, including human-in-the-loop integration. The bill does not specify the format, medium, or recipient of these disclosures, nor does it define 'program,' 'seller,' or 'distributor.' The obligation applies broadly to any AI system as defined — essentially any machine-based system that infers outputs from inputs — with no thresholds, sector limitations, or risk-level filters.
Statutory Text
Before selling or distributing a program containing artificial intelligence technology, the seller or distributor must disclose: (1) the business names of the manufacturers or creators of the AI; (2) contact information for technical experts who assist users with the AI; (3) the functions the AI performs; (4) the types of modeling the AI uses; and (5) all safety features of the AI, including but not limited to the integration of human intelligence.