,i #aaith ,,congress #b;d ,session ,h4 ,r4 #ihfb ,,in ,,the ,,house ,,of ,,representatives ,july #bb1 #bjbf,mr4 ,murphy introduced the following bill2 which was referred to the ,committee on ,veterans' ,affairs ,a ,,bill ,to direct the ,secretary of ,veterans ,affairs to seek to enter into an agreement with a federally funded research and development center for the conduct of an independent evaluation of artificial intelligence systems in use by the ,veterans ,health ,administration1 and for other purposes4 #a4 ,short title ,this ,act may be cited as the ,responsible ,artificial ,intelligence for ,veterans ,act of #bjbf4 #b4 ,independent evaluation of artificial intelligence systems used by ,veterans ,health ,administration 7a7 ,independent evaluation ,not later than #ij days after the date of the enactment of this ,act1 the ,secretary of ,veterans ,affairs shall seek to enter into an agreement with a federally funded research and development center for the conduct of an independent evaluation of artificial intelligence systems deployed1 in pilot1 or in active development1 for clinical use within the ,veterans ,health ,administration4 ,an evaluation conducted pursuant to such an agreement shall prioritize not fewer than five artificial intelligence systems that are deployed at scale or that present an elevated clinical1 operational1 or patient safety risk4 7b7 ,matters for evaluation ,an independent evaluation carried out under subsection 7a7 shall address each of the following with respect to the artificial intelligence systems evaluated3 7#a7 ,the integration and operational readiness of the systems1 including-- 7,a7 the adequacy of technological infrastructure supporting deployment and scalability2 7,b7 the interoperability of the systems with electronic health records1 medical devices1 pharmacy systems1 and other relevant platforms2 and 7,c7 the effect of the system on clinical workflows1 staffing models1 and operational processes4 7#b7 ,the governance1 monitoring1 and accountability of the systems1 including-- 7,a7 the clarity of responsibility for system outputs1 errors1 and adverse outcomes2 7,b7 policies governing consent1 data use1 secondary use1 and veteran data protections2 7,c7 lifecycle management processes1 including model updates1 retraining1 and version control2 7,d7 oversight mechanisms1 internal controls1 and audit structures2 and 7,e7 compliance with applicable ,federal privacy1 cybersecurity1 and health information laws4 7#c7 ,the performance and model integrity of the systems1 including-- 7,a7 quantitative performance metrics1 including accuracy1 sensitivity1 and specificity1 as applicable2 7,b7 the presence of disparate performance across demographic groups and the mitigation of bias2 7,c7 reliability1 repeatability1 and reproducibility across facilities and veteran populations2 7,d7 robustness to data variability and generalizability beyond training environments2 and 7,e7 transparency1 explainability1 and interpretability of model outputs sufficient to permit clinical oversight4 7#d7 ,the safety and human oversight of the systems1 including-- 7,a7 the risk of patient harm1 including failure modes and escalation pathways2 7,b7 safeguards ensuring that the artificial intelligence system functions as a clinical decision-support tool and does not supplant clinical judgment2 and 7,c7 whether clinicians retain the authority and technical ability to override or contest system outputs4 7#e7 ,the veteran-centric design and adoption of the systems1 including-- 7,a7 alignment of the system with the needs1 accessibility requirements1 and preferences of veterans2 7,b7 usability within clinical workflows2 7,c7 the availability of training resources and implementation support for clinical staff2 and 7,d7 the degree of clinician and veteran trust in system outputs4 7#f7 ,the cost and resource stewardship of the systems1 including-- 7,a7 acquisition1 deployment1 sustainment1 and cloud or compute costs2 7,b7 evidence of cost-effectiveness or cost-utility relative to clinical outcomes2 and 7,c7 resource implications for staffing1 administrative burden1 and system maintenance4 7#g7 ,the transparency and documentation of the systems1 including-- 7,a7 whether system logic1 data sources1 training inputs1 and decision pathways are sufficiently documented and auditable2 and 7,b7 whether such documentation is available for oversight review4 7#h7 ,the scalability and effect of the systems1 including-- 7,a7 the clinical effectiveness of each system in real-world practice1 including its performance across diverse veteran populations and care settings2 7,b7 whether each system improves clinical efficiency1 including reductions in administrative burden1 wait times1 or duplicative services2 7,c7 the clinical utility of each system in supporting medical decision making and improving veteran care outcomes2 and 7,d7 where applicable1 the environmental effect of deployment at scale1 including energy consumption and sustainability implications4 7c7 ,risk-,based selection criteria ,in selecting systems for evaluation pursuant to an agreement entered into under subsection 7a71 the federally funded research and development center shall prioritize artificial intelligence systems that-- 7#a7 influence diagnosis1 treatment decisions1 triage1 eligibility determinations1 or benefits adjudication2 7#b7 are deployed across multiple medical centers or impact a significant veteran population2 7#c7 are deployed at scale across facilities1 programs1 or beneficiary populations such that errors1 bias1 or system failure could reasonably affect a significant number of veterans2 7#d7 use predictive analytics1 large language models1 imaging analysis1 or automated clinical decision-support tools2 or 7#e7 present material cybersecurity1 privacy1 or patient-safety risks4 7d7 ,access to information ,the ,secretary shall ensure that a federally funded research and development center that enters into an agreement under subsection 7a7 is provided with access to relevant ,department of ,veterans ,affairs information without delay and in a manner sufficient to permit completion of the evaluation within the timeframe required under subsection 7e74 ,such information shall include1 with respect to artificial intelligence systems used by the ,veterans ,health ,administration-- 7#a7 relevant contracts1 technical documentation1 validation studies1 and model cards2 7#b7 performance data and audit logs2 7#c7 incident reports and corrective action plans2 and 7#d7 policies governing artificial intelligence governance1 data use1 and risk management4 7e7 ,reporting requirements 7#a7 ,report on findings of evaluation 7,a7 ,in general ,an agreement entered into under subsection 7a7 shall specify that1 not later than the date that is one year after the date of the agreement1 a federally funded research and development center agrees to submit to the ,secretary and to the ,committees on ,veterans' ,affairs of the ,house of ,representatives and the ,senate a report that includes each of the following3 7i7 ,findings for each evaluated system4 7ii7 ,identification of material risks to patient safety1 data security1 equity1 or clinical integrity4 7iii7 ,recommendations for corrective action1 governance improvements1 or suspension of deployment where warranted4 7iv7 ,identification of systemic gaps in the artificial intelligence governance framework of the ,department4 7,b7 ,form ,the report required by subparagraph 7,a7 shall be submitted in unclassified and unredacted form1 but may include a classified annex4 7#b7 ,corrective action plan2 ,department accountability ,not later than #abj days after receipt of a report under paragraph 7#a71 the ,secretary shall submit to the ,committees on ,veterans' ,affairs of the ,house of ,representatives and the ,senate a report that includes-- 7,a7 a response of the ,secretary addressing each finding and recommendation contained the report2 and 7,b7 a corrective action plan that includes-- 7i7 an identification of specific actions the ,department will take to address each finding2 7ii7 an assignment of responsibility to appropriate ,department officials2 7iii7 measurable milestones and deadlines for implementation2 and 7iv7 an identification of any risk that the ,secretary determines will remain unmitigated and the justification for such determination4 7#c7 ,comptroller ,general review ,not later than #ahj days after submission of the corrective action plan under paragraph 7#b71 the ,comptroller ,general of the ,united ,states shall-- 7,a7 conduct a review of-- 7i7 the independent evaluation carried out pursuant to subsection 7a72 and 7ii7 the corrective action plan submitted under paragraph 7#b7 and the initial efforts of the ,secretary to carry out the plan2 and 7,b7 provide to the ,committees on ,veterans' ,affairs of the ,house of ,representatives and the ,senate a briefing on the review that includes any recommendations of the ,comptroller ,general for improvement4 7#d7 ,committee access and transparency ,the ,secretary shall provide to the ,committees on ,veterans' ,affairs of the ,house of ,representatives and the ,senate-- 7,a7 the full independent evaluation produced pursuant to subsection 7a71 including all supporting analyses and technical appendices2 and 7,b7 any subsequent implementation updates to the corrective action plan submitted under paragraph 7#b71 without redaction1 except for information that is classified or otherwise protected by law4 7f7 ,use of existing funds ,the ,secretary shall carry out this section using amounts otherwise authorized and appropriated to the ,department4 ,no additional amounts are authorized to be appropriated to carry out this section4 7g7 ,artificial intelligence defined ,in this section1 the term artificial intelligence has the meaning given that term in section #ejjb7c7 of the ,national ,artificial ,intelligence ,initiative ,act of #bjbj 7division ,e of ,public ,law #aaf-bhc2 #acd ,stat4 #debc2 #ae ,u4,s4,c4 #idja note74