AI Cyber Defense Act

(a) In general

(1) Engagement. Not later than 180 days after the date of the enactment of this Act, the Director of the National Institute of Standards and Technology (NIST) shall seek to engage with private sector artificial intelligence stakeholders to establish the following:

(A) A definition for artificial intelligence distillation attacks.

(B) Voluntary standards and best practices for the detection and prevention of such attacks.

(2) Publication. Not later than one year after the establishment of voluntary standards and best practices under paragraph (1), the Director shall publish on the website of NIST such voluntary standards and best practices.

(b) Standards and best practices. Not later than 180 days after the date of the enactment of this Act, the Director of NIST shall establish a roundtable, to be known as the AI Cybersecurity Roundtable (in this subsection referred to as the Roundtable), to develop voluntary standards and best practices relating to the defensive use of artificial intelligence for cybersecurity purposes. The Director shall seek representation on the Roundtable of private sector artificial intelligence stakeholders and owners and operators of critical infrastructure. Not later than one year after the establishment of the Roundtable, the Director shall publish on the website of NIST such voluntary standards and best practices.

(c) Definitions. In this section:

(1) Artificial intelligence. The term artificial intelligence has the meaning given such term in section 5002 of the Artificial Intelligence Initiative Act of 2020 (15 U.S.C. 9401).

(2) Country of concern. The term country of concern means—

(A) The People’s Republic of China, including the Hong Kong and Macau Special Administrative Regions.

(B) The Russian Federation.

(C) The Islamic Republic of Iran.

(D) The Democratic People’s Republic of Korea.

(E) The Republic of Cuba.

(3) Critical infrastructure. The term critical infrastructure has the meaning given such term in section 1016(e) of Public Law 107–56 (42 U.S.C. 5195c(e)).

(4) Entity of concern

(A) In general. The term entity of concern means any foreign person or entity that—

(i) is headquartered in, or the parent company of which is headquartered in, a country of concern;

(ii) is operating under the direction or control of any entity headquartered in, or the parent company of which is headquartered in, a country of concern; or

(iii) the Director of NIST determines to be an entity of concern for purposes of this section.

(B) Exception. The term entity of concern does not include a person who—

(i) lawfully admitted for permanent residence (as such term is defined in section 101(a) of the Immigration and Nationality Act (8 U.S.C. 1101(a))); or

(ii) a citizen of the United States.

(5) Private sector artificial intelligence stakeholders. The term private sector artificial intelligence stakeholders means any corporation, partnership, limited liability company, sole proprietorship, or other business entity that—

(A) is not a unit, agency, instrumentality, or wholly owned enterprise of a Federal, State, local, or Tribal government;

(B) develops, trains, deploys, offers, sells, or licenses one or more artificial intelligence systems, or provides substantial computing infrastructure, data, or technical services for the development or deployment of such systems, as part of its ordinary business activities does so for commercial purposes, including for profit, in exchange for consideration, or in furtherance of a business enterprise, regardless of whether such entity has yet generated revenue; and

(C) is not an entity of concern.