,,ii #aahth ,,congress #b;d ,session ,s4 #dbcj ,,in ,,the ,,senate ,,of ,,the ,,united ,,states ,may #a1 #bjbd,mr4 ,warner 7for himself and ,mr4 ,tillis7 introduced the following bill2 which was read twice and referred to the ,committee on ,commerce1 ,science1 and ,transportation ,a ,,bill ,to improve the tracking and processing of security and safety incidents and risks associated with artificial intelligence1 and for other purposes4 #a4 ,short title ,this ,act may be cited as the ,secure ,artificial ,intelligence ,act of #bjbd or the ,secure ,a4,i4 ,act of #bjbd4 #b4 ,definitions ,in this ,act3 7#a7 ,artificial intelligence safety incident ,the term artificial intelligence safety incident means an event that increases the risk that operation of an artificial intelligence system will-- 7,a7 result in physical or psychological harm2 or 7,b7 lead to a state in which human life1 health1 property1 or the environment is endangered4 7#b7 ,artificial intelligence security incident ,the term artificial intelligence security incident means an event that increases-- 7,a7 the risk that operation of an artificial intelligence system occurs in a way that enables the extraction of information about the behavior or characteristics of an artificial intelligence system by a third party2 or 7,b7 the ability of a third party to manipulate an artificial intelligence system in order to subvert the confidentiality1 integrity1 or availability of an artificial intelligence system or adjacent system4 7#c7 ,artificial intelligence security vulnerability ,the term artificial intelligence security vulnerability means a weakness in an artificial intelligence system that could be exploited by a third party to subvert1 without authorization1 the confidentiality1 integrity1 or availability of an artificial intelligence system1 including through techniques such as-- 7,a7 data poisoning2 7,b7 evasion attacks2 7,c7 privacy-based attacks2 and 7,d7 abuse attacks4 7#d7 ,counter-artificial intelligence ,the term counter-artificial intelligence means techniques or procedures to extract information about the behavior or characteristics of an artificial intelligence system1 or to learn how to manipulate an artificial intelligence system1 in order to subvert the confidentiality1 integrity1 or availability of an artificial intelligence system or adjacent system4 #c4 ,voluntary tracking and processing of security and safety incidents and risks associated with artificial intelligence 7a7 ,processes and procedures for vulnerability management ,not later than #ahj days after the date of the enactment of this ,act1 the ,director of the ,national ,institute of ,standards and ,technology shall-- 7#a7 initiate a process to update processes and procedures associated with the ,national ,vulnerability ,database of the ,institute to ensure that the database and associated vulnerability management processes incorporate artificial intelligence security vulnerabilities to the greatest extent practicable2 and 7#b7 identify any characteristics of artificial intelligence security vulnerabilities that make utilization of the ,national ,vulnerability ,database inappropriate for their management and develop processes and procedures for vulnerability management for those vulnerabilities4 7b7 ,voluntary tracking of artificial intelligence security and artificial intelligence safety incidents 7#a7 ,voluntary database required ,not later than #a year after the date of the enactment of this ,act1 the ,director of the ,institute1 in coordination with the ,director of the ,cybersecurity and ,infrastructure ,security ,agency1 shall-- 7,a7 develop and establish a comprehensive1 voluntary database to publicly track artificial intelligence security and artificial intelligence safety incidents2 and 7,b7 in establishing the database under subparagraph 7,a7-- 7i7 establish mechanisms by which private sector entities1 public sector organizations1 civil society groups1 and academic researchers may voluntarily share information with the ,institute on confirmed or suspected artificial intelligence security or artificial intelligence safety incidents1 in a manner that preserves confidentiality of any affected party2 7ii7 leverage1 to the greatest extent possible1 standardized disclosure and incident description formats2 7iii7 develop processes to associate reports pertaining to the same incident with a single incident identifier2 7iv7 establish classification1 information retrieval1 and reporting mechanisms that sufficiently differentiate between artificial intelligence security incidents and artificial intelligence safety incidents2 and 7v7 create appropriate taxonomies to classify incidents based on relevant characteristics1 impact1 or other relevant criteria4 7#b7 ,identification and treatment of material artificial intelligence security or artificial intelligence safety risks 7,a7 ,in general ,upon receipt of relevant information on an artificial intelligence security or artificial intelligence safety incident1 the ,director of the ,institute shall determine whether the described incident presents a material artificial intelligence security or artificial intelligence safety risk sufficient for inclusion in the database developed and established under paragraph 7#a74 7,b7 ,priorities ,in evaluating a reported incident pursuant to paragraph 7#a71 the ,director shall prioritize inclusion in the database cases in which a described incident-- 7i7 describes an artificial intelligence system used in critical infrastructure or safety-critical systems2 7ii7 would result in a high-severity or catastrophic impact to the people or economy of the ,united ,states2 or 7iii7 includes an artificial intelligence system widely used in commercial or public sector contexts4 7#c7 ,reports and anonymity ,the ,director shall populate the voluntary database developed and established under paragraph 7#a7 with incidents based on public reports and information shared using the mechanism established pursuant to subparagraph 7,b77i7 of such paragraph1 ensuring that any incident description sufficiently anonymizes those affected1 unless those who are affected have consented to their names being included in the database4 #d4 ,updating processes and procedures relating to ,common ,vulnerabilities and ,exposures ,program and evaluation of consensus standards relating to artificial intelligence security vulnerability reporting 7a7 ,definitions ,in this section3 7#a7 ,common ,vulnerabilities and ,exposures ,program ,the term ,common ,vulnerabilities and ,exposures ,program means the reference guide and classification system for publicly known information security vulnerabilities sponsored by the ,cybersecurity and ,infrastructure ,security ,agency4 7#b7 ,relevant congressional committees ,the term relevant congressional committees means-- 7,a7 the ,committee on ,homeland ,security and ,governmental ,affairs1 the ,committee on ,commerce1 ,science1 and ,transportation1 the ,select ,committee on ,intelligence1 and the ,committee on the ,judiciary of the ,senate2 and 7,b7 the ,committee on ,oversight and ,accountability1 the ,committee on ,energy and ,commerce1 the ,permanent ,select ,committee on ,intelligence1 and the ,committee on the ,judiciary of the ,house of ,representatives4 7b7 ,in general ,not later than #ahj days after the date of enactment of this ,act1 the ,director of the ,cybersecurity and ,infrastructure ,security ,agency shall-- 7#a7 initiate a process to update processes and procedures associated with the ,common ,vulnerabilities and ,exposures ,program to ensure that the program and associated processes identify and enumerate artificial intelligence security vulnerabilities to the greatest extent practicable2 and 7#b7 identify any characteristic of artificial intelligence security vulnerabilities that make utilization of the ,common ,vulnerabilities and ,exposures ,program inappropriate for their management and develop processes and procedures for vulnerability identification and enumeration for those artificial intelligence security vulnerabilities4 7c7 ,evaluation of consensus standards 7#a7 ,in general ,not later than #cj days after the date of enactment of this ,act1 the ,director of the ,national ,institute of ,standards and ,technology shall initiate a multi-stakeholder process to evaluate whether existing voluntary consensus standards for vulnerability reporting effectively accommodate artificial intelligence security vulnerabilities4 7#b7 ,report 7,a7 ,submission ,not later than #ahj days after the date on which the evaluation under paragraph 7#a7 is carried out1 the ,director shall submit a report to the relevant congressional committees on the sufficiency of existing vulnerability reporting processes and standards to accommodate artificial intelligence security vulnerabilities4 7,b7 ,post-report action ,if the ,director concludes in the report submitted under subparagraph 7,a7 that existing processes do not sufficiently accommodate reporting of artificial intelligence security vulnerabilities1 the ,director shall initiate a process1 in consultation with the ,director of the ,national ,institute of ,standards and ,technology and the ,director of the ,office of ,management and ,budget1 to update relevant vulnerability reporting processes1 including the ,department of ,homeland ,security ,binding ,operational ,directive #bj-ja1 or any subsequent directive4 7d7 ,best practices ,not later than #ij days after the date of enactment of this ,act1 the ,director of the ,cybersecurity and ,infrastructure ,security ,agency shall1 in collaboration with the ,director of the ,national ,security ,agency and the ,director of the ,national ,institute of ,standards and ,technology and by leveraging efforts of the ,information ,communications ,technology ,supply ,chain ,risk ,management ,task ,force to the greatest extent practicable1 convene a multi-stakeholder process to encourage the development and adoption of best practices relating to addressing supply chain risks associated with training and maintaining artificial intelligence models1 which shall ensure consideration of supply chain risks associated with-- 7#a7 data collection1 cleaning1 and labeling1 particularly the supply chain risks of reliance on remote workforce and foreign labor for such tasks2 7#b7 inadequate documentation of training data and test data storage1 as well as limited provenance of training data2 7#c7 human feedback systems used to refine artificial intelligence systems1 particularly the supply chain risks of reliance on remote workforce and foreign labor for such tasks2 7#d7 the use of large-scale1 open-source datasets1 particularly the supply chain risks to repositories that host such datasets for use by public and private sector developers in the ,united ,states2 and 7#e7 the use of proprietary datasets containing sensitive or personally identifiable information4 7e7 ,rule of construction ,to the extent practicable1 the ,director shall examine the reporting requirements pursuant to division ,y of the ,cyber ,incident ,reporting for ,critical ,infrastructure ,act of #bjbb 7,public ,law #aag-ajc7 and the amendments made by that division and ensure that the requirements under this section are not duplicative of requirements set forth in that division and the amendments made by that division4 #e4 ,establishment of ,artificial ,intelligence ,security ,center 7a7 ,establishment ,not later than #ij days after the date of the enactment of this ,act1 the ,director of the ,national ,security ,agency shall establish an ,artificial ,intelligence ,security ,center within the ,cybersecurity ,collaboration ,center of the ,national ,security ,agency4 7b7 ,functions ,the functions of the ,artificial ,intelligence ,security ,center shall be as follows3 7#a7 ,making available a research test-bed to private sector and academic researchers1 on a subsidized basis1 to engage in artificial intelligence security research1 including through the secure provision of access in a secure environment to proprietary third-party models with the consent of the vendors of the models4 7#b7 ,developing guidance to prevent or mitigate counter-artificial intelligence techniques4 7#c7 ,promoting secure artificial intelligence adoption practices for managers of national security systems 7as defined in section #ceeb of title #dd1 ,united ,states ,code7 and elements of the defense industrial base4 7#d7 ,coordinating with the ,artificial ,intelligence ,safety ,institute within the ,national ,institute of ,standards and ,technology4 7#e7 ,such other functions as the ,director considers appropriate4 7c7 ,test-,bed requirements 7#a7 ,access and terms of usage 7,a7 ,researcher access ,the ,director shall establish terms of usage governing researcher access to the test-bed made available under subsection 7b77#a71 with limitations on researcher publication only to the extent necessary to protect classified information or proprietary information concerning third-party models provided through the consent of model vendors4 7,b7 ,availability to ,federal agencies ,the ,director shall ensure that the test-bed made available under subsection 7b77#a7 is also made available to other ,federal agencies on a cost-recovery basis4 7#b7 ,use of certain infrastructure and other resources ,in carrying out subsection 7b77#a71 the ,director shall leverage1 to the greatest extent practicable1 infrastructure and other resources provided under section #e4b of the ,executive ,order dated ,october #cj1 #bjbc 7relating to safe1 secure1 and trustworthy development and use of artificial intelligence74 7d7 ,access to proprietary models ,in carrying out this section1 ,the ,director shall establish such mechanisms as the ,director considers appropriate1 including potential contractual incentives1 to ensure the provision of access to proprietary models by qualified independent1 third-party researchers1 provided that commercial model vendors have voluntarily provided models and associated resources for such testing4