,,ii #aahth ,,congress #b;d ,session ,s4 #eeci ,,in ,,the ,,senate ,,of ,,the ,,united ,,states ,december #af1 #bjbd,mr4 ,welch 7for himself and ,mr4 ,lujn7 introduced the following bill2 which was read twice and referred to the ,committee on ,homeland ,security and ,governmental ,affairs ,a ,,bill ,to require systematic review of artificial intelligence systems before deployment by the ,federal ,government1 and for other purposes4 #a4 ,short title ,this ,act may be cited as the ,trustworthy ,by ,design ,artificial ,intelligence ,act of #bjbd or the ,,tbd ,,ai ,act of #bjbd4 #b4 ,definitions ,in this ,act3 7#a7 ,artificial intelligence system ,the term artificial intelligence system means a machine-based system that can1 for a given set of machine-defined or human-defined objectives1 make or inform predictions1 recommendations1 or decisions influencing real or virtual environments4 ,artificial intelligence systems use machine and human-based inputs to-- 7,a7 perceive real and virtual environments2 7,b7 abstract such perceptions into models through analysis in an automated manner2 and 7,c7 use model inference to formulate options for information or action4 7#b7 ,director ,the term ,director means the ,director of the ,national ,institute of ,standards and ,technology4 7#c7 ,federal agency ,the term ,federal agency means any ,federal department1 agency1 or organization4 #c4 ,guidelines for evaluation of trustworthiness of artificial intelligence systems 7a7 ,development required 7#a7 ,in general ,not later than #a year after the date of the enactment of this ,act1 the ,director shall develop and release a set of guidelines for evaluation of the trustworthiness of artificial intelligence systems4 7#b7 ,use of existing guidelines or elements 7,a7 ,in general ,in carrying out paragraph 7#a71 the ,director may use existing guidelines1 best practices1 or elements of guidelines or best practices that the ,director identifies from other sources4 7,b7 ,annotation ,if the ,director uses guidelines1 best practices1 or elements of guidelines or best practices pursuant to subparagraph 7,a7 of this paragraph in order to carry out paragraph 7#a71 the ,director shall clearly annotate such use in a central location and make such annotations available online to the public4 7#c7 ,periodic updates ,the ,director shall1 on a periodic basis but not less frequently than annually1 update the guidelines developed pursuant to paragraph 7#a74 7b7 ,components covered ,the guidelines developed pursuant to subsection 7a7 for evaluation of artificial intelligence systems shall cover the following3 7#a7 ,the models used for the artificial intelligence systems4 7#b7 ,the data used and activities conducted in training the artificial intelligence systems1 including the collection and filtering of data4 7#c7 ,the processes and techniques applied after initial training to enhance the capabilities of the artificial intelligence systems1 such as fine-tuning1 reinforcement learning1 and other post-training optimization methods4 7#d7 ,content generated by the artificial intelligence systems4 7#e7 ,the hardware systems used by the artificial intelligence systems4 7#f7 ,interactions between humans and the artificial intelligence systems that are expected to arise during intended or reasonably foreseeable use of the artificial intelligence systems4 7#g7 ,risks presented by anthropomorphic artificial intelligence systems4 7c7 ,elements of trustworthiness 7#a7 ,in general ,the guidelines developed pursuant to subsection 7a7 shall cover trustworthiness with respect to the following3 7,a7 ,validity and reliability4 7,b7 ,safety4 7,c7 ,security4 7,d7 ,resiliency4 7,e7 ,transparency and accountability4 7,f7 ,explainability and interpretability4 7,g7 ,privacy4 7,h7 ,fairness and bias4 7,i7 ,such other matters relating to safety1 security1 or trustworthiness as the ,director considers appropriate4 7#b7 ,protected classes ,the guidelines developed pursuant to subsection 7a7 shall specifically highlight and consider accuracy and bias risks relating to protected classes under ,federal law4 7d7 ,applicability ,the ,director shall ensure that the guidelines developed pursuant to subsection 7a7 are developed so that they include an assessment of the trustworthiness of all components covered under subsection 7b7 with regards to all elements under subsection 7c71 but account for circumstances in which certain assessment methods or recommendations may not be applicable to certain components or elements may not be applicable4 7e7 ,limitation relating to synthetic content ,under the guidelines developed pursuant to subsection 7a71 the ,director shall identify appropriate mechanisms to manage the risks from relying on synthetic content or content created by artificial intelligence systems to improve a dataset or model1 or to meet evaluation guidelines4 7f7 ,developing robust guidelines ,the ,director shall ensure that the guidelines developed pursuant to subsection 7a7 are developed in such a manner that encourages transparency1 cooperation1 and collaboration with developers or evaluators of artificial intelligence systems1 academia1 and civil society sufficient to independently verify the elements set forth under subsection 7c77#a74 7g7 ,iterative evaluation ,the ,director shall ensure that the guidelines developed pursuant to subsection 7a7 cover how best to evaluate the trustworthiness of artificial intelligence systems iteratively1 throughout the design1 development1 and deployment lifecycle of an artificial intelligence system4 7h7 ,report to ,congress 7#a7 ,in general ,the ,director shall submit to ,congress a report on any expected barriers to implementing and adhering to the guidelines developed pursuant to subsection 7a71 especially with respect to transparency1 cooperation1 or collaboration barriers with developers of artificial intelligence systems4 7#b7 ,form ,the report under paragraph 7#a7 shall be submitted in unclassified form1 but may include a classified appendix1 if necessary4 #d4 ,federal deployment of artificial intelligence systems 7a7 ,covered use defined ,in this section1 with respect to an artificial intelligence system1 the term covered use-- 7#a7 means use in any automated decision making2 and 7#b7 does not include any use that the ,director exempts from any portion of the guidelines issued under section #c7;a71 including any use that-- 7,a7 is subject to evaluation by existing national security assessments2 or 7,b7 is an edge case1 especially a time sensitive or emergency case1 as determined by the ,director4 7b7 ,existing artificial intelligence systems ,with respect to any artificial intelligence system deployed for a covered use that is in use by a ,federal agency before the date of enactment of this ,act1 the ,federal agency shall evaluate the artificial intelligence system deployment to ensure the artificial intelligence system meets the guidelines developed under section #c7;a7 not later than #b years after the effective date of this section or cease using the artificial intelligence system4 7c7 ,new artificial intelligence systems ,the head of each ,federal agency shall ensure that each new acquisition or integration of an artificial intelligence system by the ,federal agency for a covered use meets the guidelines developed under section #c7;a7 prior to deployment of the artificial intelligence system4 7d7 ,labeling 7#a7 ,in general ,the head of each ,federal agency shall identify as compliant1 and make publicly available1 in accordance with all applicable classification requirements and national security restrictions1 the documentation of evaluation status and compliance details for1 each artificial intelligence system deployment for a covered use by the ,federal agency that meets the guidelines developed under section #c7;a74 7#b7 ,noncompliant deployments ,not later than #b years after the effective date of this section1 the head of each ,federal agency that deploys an artificial intelligence system that is not compliant or is not evaluated pursuant to paragraph 7#a71 shall-- 7,a7 make publicly available documentation of each such deployment2 and 7,b7 with respect to each such deployment1 report-- 7i7 the status of the evaluation2 7ii7 progress made towards compliance2 7iii7 a clear1 specific justification for any delay2 and 7iv7 any barriers to compliance1 including resource constraints4 7#c7 ,report to ,congress 7,a7 ,in general ,not later than #c years after the effective date of this section1 the head of each ,federal agency shall submit to ,congress a report of each deployment described in paragraph 7#b74 7,b7 ,form ,the report under subparagraph 7,a7 shall be submitted in unclassified form1 but may include a classified appendix1 if necessary4 7e7 ,chief ,,ai ,officers 7#a7 ,in general ,not later than #abj days after the effective date of this section1 the head of each ,federal agency shall designate a ,chief ,artificial ,intelligence ,officer1 with responsibility for the management1 governance1 acquisition1 and oversight processes of the ,federal agency relating to artificial intelligence1 including implementation of the guidelines developed under section #c7;a74 7#b7 ,full-time employee ,to the extent practicable1 each ,chief ,artificial ,intelligence ,officer designated under paragraph 7#a7 shall be a full-time employee of the ,federal agency on the date of the designation4 7#c7 ,seniority ,with respect to the ,chief ,artificial ,intelligence ,officer of any agency described in section #ija7;b7 of title #ca1 ,united ,states ,code1 the ,chief ,artificial ,intelligence ,officer shall be an executive with a position classified above ,,gs-#ae of the ,general ,schedule or the equivalent4 7f7 ,effective date ,this section shall take effect on the date that the guidelines developed pursuant to section #c7;a7 are released4