Signed by Governor JB Pritzker

Vote Smart's Synopsis:

Vote to pass a bill that establishes new safety requirements for artificial intelligence developers in Illinois.

Highlights:

• Defines “foundation model” as an artificial intelligence (AI) model that is (Sec. 5):

• Trained on a broad data set;

• Designed for generality of output; and

• Adaptable to a wide range of distinctive tasks.

• Defines “frontier AI framework” as documented technical and organizational protocols to manage, assess, and mitigate catastrophic risks (Sec. 5).

• Defines “frontier model” as a foundation model that was trained using a quantity of computing power greater than 1026 integer or floating-point operations for either the original training run or any subsequent training (Sec. 5).

• Defines “frontier developer” as a person who trains a frontier model using computing power that meets the specifications under the definition of “frontier model” (Sec. 5).

• Defines “large frontier developer” as a frontier developer that, together with its affiliates, collectively had annual gross revenues over $500 million in the preceding calendar year (Sec. 5).

• Defines “catastrophic risk” as a foreseeable and material risk that a frontier developer’s development, storage, use, or deployment of a frontier model will materially contribute to the death or serious injury of more than 50 people, or more than $1 billion of property damage (Sec. 5).

• Specifies that a catastrophic risk may include a frontier model doing any of the following (Sec. 5):

• Providing expert-level assistance in the creation or release of a chemical, biological, radiological, or nuclear weapon;

• Engaging in conduct with no meaningful human oversight, intervention, or supervision that constitutes a cyberattack, murder, assault, extortion, or theft; or

• Evading the control of its frontier developer or user.

• Requires large frontier developers to publicly publish on their website a framework detailing how they manage, assess, and mitigate potential catastrophic risks (Sec. 10).

• Requires large frontier developers’ risk mitigation frameworks to be updated at least once per year (Sec. 10).

• Requires large frontier developers to publish a transparency report alongside a new or substantially modified frontier model that includes the following (Sec. 10):

• The website of the developer;

• A mechanism that allows an individual to communicate with the developer;

• The release date of the model;

• The languages supported by the model;

• The modalities of output supported by the model;

• The intended uses of the model;

• Any restrictions or conditions on uses of the model;

• Assessments of catastrophic risks from the model; and

• The extent to which third-party evaluators were involved.

• Requires large frontier developers to annually retain a third party to perform an independent audit of compliance with the requirements of this section (Sec. 10).

• Requires large frontier developers to submit a summary of internal catastrophic risk assessments to the Illinois Emergency Management Agency and Office of Homeland Security every three months (Sec. 10).

• Prohibits large frontier developers from making false or misleading statements regarding catastrophic risks or compliance with their AI frameworks (Sec. 10).

• Requires large frontier developers to disclose certain information such as the names under which they conduct business, the principal address of business, the names of individuals with a 5% or greater interest in the developer, and contact information for the developer (Sec. 18).

• Prohibits a frontier developer from retaliating against whistleblowers (Sec. 20).

• Establishes civil penalties for large frontier developers that fail to publish or submit documents required under this act (Sec. 25).

• Specifies that this Act takes effect January 1, 2027 (Sec. 99).

Full Bill Text

Vote Smart's Synopsis:

Vote to pass a bill that establishes new safety requirements for artificial intelligence developers in Illinois.

Highlights:

• Defines “foundation model” as an artificial intelligence (AI) model that is (Sec. 5):

• Trained on a broad data set;

• Designed for generality of output; and

• Adaptable to a wide range of distinctive tasks.

• Defines “frontier AI framework” as documented technical and organizational protocols to manage, assess, and mitigate catastrophic risks (Sec. 5).

• Defines “frontier model” as a foundation model that was trained using a quantity of computing power greater than 1026 integer or floating-point operations for either the original training run or any subsequent training (Sec. 5).

• Defines “frontier developer” as a person who trains a frontier model using computing power that meets the specifications under the definition of “frontier model” (Sec. 5).

• Defines “large frontier developer” as a frontier developer that, together with its affiliates, collectively had annual gross revenues over $500 million in the preceding calendar year (Sec. 5).

• Defines “catastrophic risk” as a foreseeable and material risk that a frontier developer’s development, storage, use, or deployment of a frontier model will materially contribute to the death or serious injury of more than 50 people, or more than $1 billion of property damage (Sec. 5).

• Specifies that a catastrophic risk may include a frontier model doing any of the following (Sec. 5):

• Providing expert-level assistance in the creation or release of a chemical, biological, radiological, or nuclear weapon;

• Engaging in conduct with no meaningful human oversight, intervention, or supervision that constitutes a cyberattack, murder, assault, extortion, or theft; or

• Evading the control of its frontier developer or user.

• Requires large frontier developers to publicly publish on their website a framework detailing how they manage, assess, and mitigate potential catastrophic risks (Sec. 10).

• Requires large frontier developers’ risk mitigation frameworks to be updated at least once per year (Sec. 10).

• Requires large frontier developers to publish a transparency report alongside a new or substantially modified frontier model that includes the following (Sec. 10):

• The website of the developer;

• A mechanism that allows an individual to communicate with the developer;

• The release date of the model;

• The languages supported by the model;

• The modalities of output supported by the model;

• The intended uses of the model;

• Any restrictions or conditions on uses of the model;

• Assessments of catastrophic risks from the model; and

• The extent to which third-party evaluators were involved.

• Requires large frontier developers to annually retain a third party to perform an independent audit of compliance with the requirements of this section (Sec. 10).

• Requires large frontier developers to submit a summary of internal catastrophic risk assessments to the Illinois Emergency Management Agency and Office of Homeland Security every three months (Sec. 10).

• Prohibits large frontier developers from making false or misleading statements regarding catastrophic risks or compliance with their AI frameworks (Sec. 10).

• Requires large frontier developers to disclose certain information such as the names under which they conduct business, the principal address of business, the names of individuals with a 5% or greater interest in the developer, and contact information for the developer (Sec. 18).

• Prohibits a frontier developer from retaliating against whistleblowers (Sec. 20).

• Establishes civil penalties for large frontier developers that fail to publish or submit documents required under this act (Sec. 25).

• Specifies that this Act takes effect January 1, 2027 (Sec. 99).

Full Bill Text