AI System Risk and Transparency Act

 

 

This act shall establish a basic, risk based framework for the regulation of AI systems in the State of Minnesota to

ensure protection of individual rights and safety.

 

 

The potential risks associated with AI include potential for bias and discrimination. As such, there are presently no

comprehensive regulatory laws governing the use of AI in Minnesota. This act will:

1. Protect civil rights in areas including employment, consumer lending and access to government services.

2. Ban from operation all high risk AI systems which have the potential to harm the public or individuals within the

state.

3. Mandate that providers disclose to consumers where AI technology is being utilized.

4. Foster a safe, low risk environment in which AI innovations may flourish.

 

For purposes of this act:

a. AI SYSTEM:means a computer system or network that generates or makes decisions based upon data or information.

b. HIGHRISK AI SYSTEM:includes any AI system that is utilized in connection with the provision of governmental services,

employment, credit or law enforcement.

c. UNACCEPTABLE RISK AI SYSTEM:means any AI system whose operational impact upon the general public and/or individual

citizens could cause serious harm or risk to life, liberty, property or personal rights.

d. GENERATIVE AI:means any type of AI system capable of creating original content.

e. "PROVIDER:means any entity responsible for the development, deployment, marketing, sale or distribution of an AI

system.

 

 

 

SECTION IV — FUNDING

A total amount of $500,000 shall be allocated to the Office of the Attorney General to employ an attorney with

experience in AI regulation and compliance to serve as an AI Enforcement Analyst. In addition, these funds will be used

to develop public education materials and to assist counties in their efforts to enforce this statute.

 

 

 

SECTION V — PENALITIES

A. Any provider who employs an UNACCEPTABLE RISK AI SYSTEM in violation of Section II(b) above shall be subject to

penalties of fines up to $10 million or five percent of its worldwide gross revenues for each day during which the

violation occurred.

B. Any provider who employs a HIGH RISK AI SYSTEM in violation of Section II(c) above shall be subject to penalties of

fines up to $1 million or two percent of its worldwide gross revenues for each day during which the violation occurred.

C. Any provider who fails to comply with the requirements of Section III(d) above shall be subject to penalties of fines

up to $10,000 per day for each day during which the violation continued. Such violations shall be treated as deceptive

trade practices under applicable statutes.

D. The Attorney General and county attorneys shall have jurisdiction to enforce the provisions of this act.

 

SECTION VI — EFFECTIVE DATE

This bill shall take effect on January 1, 2027