,i #aaith ,,congress #b;d ,session ,h4 ,r4 #ajcfb ,,in ,,the ,,house ,,of ,,representatives ,september #ad1 #bjbf,mr4 ,gottheimer 7for himself and ,mr4 ,lawler7 introduced the following bill2 which was referred to the ,committee on ,science1 ,space1 and ,technology1 and in addition to the ,committee on ,oversight and ,government ,reform1 for a period to be subsequently determined by the ,speaker1 in each case for consideration of such provisions as fall within the jurisdiction of the committee concerned ,a ,,bill ,to provide for certain artificial intelligence agent discovery and security standards1 and for other purposes4 #a4 ,short title ,this ,act may be cited as the ,stop ,rogue ,,ai ,act4 #b4 ,,ai agent discovery and security standards 7a7 ,,ai agent discovery and security standards 7#a7 ,in general ,not later than #a year after the date of the enactment of this ,act and annually thereafter1 the ,director of the ,national ,institute of ,standards and ,technology 7in this section referred to as the ,director71 in coordination with the ,assistant ,secretary of ,commerce for ,communications and ,information 7in this section referred to as the ,assistant ,secretary71 shall develop1 publish1 and maintain standards1 guidelines1 and best practices for the secure development1 deployment1 and operation of artificial intelligence agents by an organization4 ,such standards1 guidelines1 and best practices shall-- 7,a7 maintain the capability to continuously discover1 inventory1 verify1 and maintain organizational control over all ,,ai agents operating within or interacting with the information systems1 networks1 applications1 services1 or digital environments of such organizations2 7,b7 maintain organizational control over ,,ai agents deployed by such organization2 7,c7 integrate discovery mechanisms into broader cybersecurity and risk management processes1 consistent with defense-in-depth2 7,d7 apply discovery and verification controls consistently across ,,ai agents regardless of whether such ,,ai agents are developed internally1 acquired from third-party vendors1 or operated through external services2 7,e7 evaluate the security1 safety1 correctness1 and reliability of ,,ai agents before such ,,ai agents are deployed by such organization and continuously after such deployment2 7,f7 enable continuous runtime monitoring and1 where appropriate1 inline detection and interception of ,,ai agent interactions with tools1 data sources1 information systems1 and other ,,ai agents1 including detection of prompt injection1 data exfiltration1 anomalous tool invocation1 and behavioral drift from an ,,ai agent's approved operational baseline2 7,g7 implement cryptographically verifiable provenance mechanisms sufficient to identify the entity responsible for creating or operating an ,,ai agent2 and 7,h7 generate and retain tamper-evident1 standardized logs of material ,,ai agent actions1 and ensure such logs are portable and accessible1 as appropriate and consistent with law1 to deploying organizations and authorized relying parties4 7#b7 ,,ai agent discovery as a component of cybersecurity ,in carrying out paragraph 7#a71 the ,director1 in coordination with the ,assistant ,secretary1 shall include in the standards1 guidelines1 and best practices described in such paragraph ,,ai agent discovery as a necessary component of effective cybersecurity within applicable frameworks1 profiles1 and reference materials4 7#c7 ,open and interoperable discovery standards ,the ,director1 in coordination with the ,assistant ,secretary1 shall support the development and adoption of open1 vendor-agnostic1 and interoperable standards for ,,ai agent discovery mechanisms4 ,in carrying out this paragraph1 the ,director and the ,assistant ,secretary shall-- 7,a7 promote the use of existing internet infrastructure and identity-based discovery mechanisms1 including domain name system-based approaches1 cryptographically verifiable ,,ai agent identity and registry frameworks1 or functionally equivalent mechanisms1 to enable secure and scalable ,,ai agent discovery and ,,ai agent identity verification2 7,b7 ensure that such standards are globally interoperable1 distributed1 and not dependent on proprietary or platform-specific registries2 and 7,c7 engage with multistakeholder processes1 including industry1 civil society1 and technical standards bodies1 to support broad adoption and international coordination4 7#d7 ,minimum organizational requirements ,standards1 guidelines1 and best practices developed under this subsection shall provide that organizations deploying ,,ai agents-- 7,a7 maintain a continuous1 machine-readable inventory of all ,,ai agents1 using standardized1 vendor-agnostic naming conventions2 7,b7 implement ,,ai agent identity verification and trust verification mechanisms that are independent and cryptographically verifiable at both the network and application layers2 and 7,c7 do not rely solely on self-attested or single-provider assertions for establishing ,,ai agent identity4 7#e7 ,guidance1 coordination1 and demonstration projects ,the ,director1 in coordination with the ,assistant ,secretary1 shall-- 7,a7 incorporate the standards1 guidelines1 and best practices developed under this subsection into existing ,national ,institute of ,standards and ,technology frameworks and guidance2 7,b7 ensure that standards1 guidelines1 and best practices developed under this subsection are consistent with and not duplicative of other efforts by ,national ,institute of ,standards and ,technology to establish standards related to ,,ai agents2 7,c7 conduct or support demonstration projects1 including through the ,national ,cybersecurity ,center of ,excellence1 to evaluate the effectiveness of open agent discovery mechanisms2 7,d7 coordinate with the ,administrator of the ,national ,telecommunications and ,information ,administration to promote outreach and adoption through multistakeholder processes and international engagement1 as appropriate2 and 7,e7 coordinate with the ,director of the ,cybersecurity and ,infrastructure ,security ,agency of the ,department of ,homeland ,security to ensure ,,ai agent discovery standards are reflected in applicable ,federal civilian agency security guidance and binding operational directives1 as appropriate4 7b7 ,federal procurement requirements for ,,ai agent security 7#a7 ,in general ,not later than #ah months after the publication of standards1 guidelines1 and best practices under subsection 7a7 and annually thereafter1 the ,federal ,acquisition ,regulatory ,council shall propose revisions to the ,federal ,acquisition ,regulation to require contractors and ,federal agencies procuring or deploying1 as the case may be1 ,,ai agents or information systems that interact with ,,ai agents to comply with such standards1 guidelines1 and best practices4 7#b7 ,required contract elements ,revisions proposed under paragraph 7#a7 shall ensure contracts for the procurement or deployment of ,,ai agents include requirements that the contractor-- 7,a7 maintain a continuous1 machine-readable inventory of all ,,ai agents deployed under such contract1 using standardized1 vendor-agnostic naming conventions consistent with standards1 guidelines1 and best practices developed under subsection 7a72 7,b7 implement ,,ai agent identity verification mechanisms that are cryptographically verifiable at both the network and application layers2 7,c7 ensure ,,ai agent discovery and verification capabilities are accessible to the ,federal agency that has entered into such a contract without reliance on the contractor with which such ,federal agency has so entered into such a contract2 7,d7 enable the ,federal agency that has entered into such a contract to exercise organizational control over ,,ai agent activity1 including the ability to allow1 deny1 or constrain ,,ai agent-to-,,ai agent and ,,ai agent-to-information system interactions2 7,e7 enable the ,federal agency that has entered into such a contract to provide cryptographically verifiable provenance information to authorized entities interacting with the ,,ai agent1 consistent with the policies of such ,federal agency2 and 7,f7 generate tamper-evident1 standardized logs of material ,,ai agent actions1 and ensure such logs are portable and accessible to the ,federal agency that has entered into such a contract and1 where appropriate and consistent with law and such contract1 authorized relying parties4 7#c7 ,required contract elements for information systems interacting with ,,ai agents ,revisions proposed under paragraph 7#a7 shall ensure contracts for the procurement or deployment of information systems that ,,ai agents interact with include requirements that the contractor-- 7,a7 maintain a continuous1 machine-readable inventory of all ,,ai agents that interact with such an information system1 using standardized1 vendor-agnostic naming conventions consistent with standards1 guidelines1 and best practices developed under subsection 7a72 7,b7 implement identity verification mechanisms that are cryptographically verifiable for all ,,ai agents that interact with such an information system2 7,c7 implement provenance verification mechanisms that are cryptographically verifiable for all ,,ai agents that interact with such an information system2 7,d7 ensure ,,ai agent discovery and verification capabilities are accessible to the ,federal agency that has entered into such a contract without reliance on the contractor with which such ,federal agency has so entered into such a contract2 7,e7 enable the ,federal agency that has entered into such a contract to exercise organizational control over ,,ai agents' ability to interact with such an information system1 including the ability to allow1 deny1 or constrain ,,ai agent-to-,,ai agent interactions within such an information system2 and 7,f7 generate tamper-evident1 standardized logs of material ,,ai agent actions within such an information system and ensure such logs are portable and accessible to the ,federal agency that has entered into such a contract and1 where appropriate and consistent with law and such contract1 authorized relying parties4 7#d7 ,saving provision for certain contracts ,this subsection shall not apply to contracts for the procurement or deployment1 as the case may be1 of ,,ai agents or information systems that interact with ,,ai agents entered into before the date of the enactment of this ,act4 7#e7 ,agency guidance ,not later than #ahj days after the proposal of revisions under paragraph 7#a7 and annually thereafter1 the ,director of the ,office of ,management and ,budget1 in coordination with the ,director of the ,cybersecurity and ,infrastructure ,security ,agency1 shall issue guidance to ,federal agencies regarding the following3 7,a7 ,the implementation of procurement requirements under this subsection1 including for ,,ai agents deployed through cloud services1 platform integrations1 or third-party managed environments4 7,b7 ,the effective usage of ,,ai agents by such ,federal agencies in accordance with the standards1 guidelines1 and best practices under subsection 7a74 7c7 ,definitions ,in this section3 7#a7 ,,ai agent discovery ,the term ,,ai agent discovery means the technical and organizational capability to identify1 enumerate1 verify1 and maintain a current inventory of ,,ai agents operating within1 communicating with1 or seeking access to an information system1 network1 application1 service1 or digital environment4 7#b7 ,,ai2 ,artificial intelligence ,the terms ,,ai and artificial intelligence have the meaning given the term artificial intelligence in section #ejjb of the ,national ,artificial ,intelligence ,initiative ,act of #bjbj 7#ae ,u4,s4,c4 #idja74 7#c7 ,,ai model ,the term ,,ai model means a software component of an information system that implements artificial intelligence technology and uses computational1 statistical1 or machine-learning techniques to produce outputs from a defined set of inputs4 7#d7 ,artificial intelligence agent2 ,,ai agent ,the terms artificial intelligence agent and ,,ai agent mean a software-based system that-- 7,a7 uses an ,,ai model to perceive1 plan1 or make decisions2 7,b7 autonomously interacts with other software systems1 digital services1 users1 external environments1 or ,,ai agents on behalf of a person or organization2 and 7,c7 involves minimal or no human interaction beyond its initial direction4 7#e7 ,defense-in-depth ,the term defense-in-depth means the protection of information systems by using multiple security measures1 including policies1 procedures1 and physical security1 such as antivirus software1 firewalls1 anti-spyware tools1 strong password policies1 intrusion detection systems1 biometric verification1 encryption1 and multi-factor authentication1 to reduce the risk of unauthorized access1 data breach1 or other successful attack4 7#f7 ,information system ,the term information system has the meaning given such term in section #cejb of title #dd1 ,united ,states ,code4 7#g7 ,organizational control ,the term organizational control means the technical and organizational ability to-- 7,a7 allow1 deny1 or restrict-- 7i7 an ,,ai agent's access to specific data2 7ii7 the actions an ,,ai agent can perform2 and 7iii7 the tools1 information systems1 and ,,ai agents which an ,,ai agent can use or interact with2 and 7,b7 revoke or change at any time any of the allowances1 denials1 or restrictions described in clauses 7i7 through 7iii7 of subparagraph 7,a74