美國國務(wù)院-人工智能戰(zhàn)略（英）

E

N

T

E

R

P

R

I

S

EARTIFICIALINTELLIGENCES

T

R

AT

E

G

Y

F

Y

2

0

2

4

-

F

Y

2

0

2

5Empowering

Diplomacy

through

Responsible

AIU.S.Departmentof

StateOctober

2023Quotefrom

S“[TheDepartment

wants]toensurethat

asAIand

relatedtechnologiestransform

howwelive,

how

wework,how

wecompete,how

wedefendourselves,thatwe’restayingahead

ofchange,indeedthatweare

shaping

change

and,

critically,makingsure

it

deliversforourpeople.”-

S

E

C

R

E

T

AR

Y

AN

T

O

N

Y

J

.

BL

I

N

K

E

NTABLE

ofCONTENTS45Letterfrom

theSecretary

of

StateExecutive

Summary6Introduction7Strategic

Drivers8Departmentof

State

AIVisionStrategic

Goals

and

Objectives99GOAL

1:

LeverageSecure

AIInfrastructure10111213GOAL

2:

Foster

a

Culture

that

Embraces

AITechnologyGOAL

3:

Ensure

AIis

AppliedResponsiblyGOAL

4:

InnovateMoving

ForwardU.S.DepartmentofStateEnterpriseAIStrategyPage3ENTERPRISE

AI

STRATEGYEmpowering

Diplomacy

throughResponsible

AIAletterfrom

theSecretary

of

StateThroughout

itshistory,theStateDepartmenthasbeeninanear-constantstateofrenewal,strivingtoensurethatAmericandiplomacy

wasservingournationalinterests

inarapidlychanging

andincreasinglycomplexworld.New

challengesmeanthattheDepartmentmustadapt

andtransform,

andIamcommittedtoensuringourdiplomatshavethebesttools,tradecraft,andtechnologies

toservetheAmerican

peopleathomeandabroad.AI,

especially

generative

AI,

serves

asatransformativetool

inourdiplomaticarsenal.Fromcitizenservicestoforeignpolicyanalysisandeven

negotiation

advantages,AI

offersusanopportunitytoenhanceoureffortswithoriginalinsightsandbeyond-humanprocessingspeed.AstheUnitedStatesworkstoadvanceavisionofeffective,ethical,andresponsibleuseofAI

globally,itisimportantweleadbyexampleinouruseofAIhereintheDepartment.HarnessingthebenefitsofAItoadvanceourforeignpolicyandincreasemanagementefficiencyintheDepartmentrequiresasecureandAI-readytechnologicalinfrastructure;therecruitment,upskilling,

andretention

ofanAI-readyworkforce;consistent,responsiblegovernanceandstandards;

andtangibledeploymentofAI

toimproveouroperations.

Byfollowingtheprinciplesoutlinedinthisstrategy,theDepartmentwillbestpositionourdiplomatstotakeadvantageoftheforeignpolicyinsightsandoperational

efficienciesAIpromises,whilemitigatingitsrisks.Iextendmygratitudetoallthosewhoplayedaroleinshaping

thisstrategy,andwhowillworktomakeitareality

inthecomingyears.The

Department’sEnterpriseGovernanceBoard,AISteering

Committee,andResponsibleAI

Official,

amongmanyothers,willworkdiligentlytoadvanceethicalandeffectiveAI

intheDepartment,complywithlegislativeandpolicyrequirements,andactively

innovatenewtechnologicalandpolicyapproachestogovernAIasthetechnology

itselfcontinuestochange.Collaborationremainsattheheartofoursuccess.

The

synergybetween

ourbureaus,

ourincredibleworkforce,andourinteragency

partnerswilldetermineoursuccess

inincorporatingresponsibleAIintoourdiplomaticefforts.Ourmissionremainsclear:tochampion

theinterests,safety,andeconomicprosperityoftheAmericanpeople.

WithyourcontinueddedicationandthepowerofinnovativeAIsolutions,

Ihavenodoubt

thattheDepartmentofStatewillremainaleaderinsafe,secure,andtrustworthyAI.ANT

ONY

J.

BLINKENU.S.

SecretaryofState<<

RETURN

TO

TABLE

OF

CONTENTSU.S.DepartmentofStateEnterpriseAIStrategyPage4E

X

E

C

U

T

I

V

E

S

U

M

M

A

R

YE

A

I

SVISIONTheDepartment

of

State

will

responsiblyand

securely

harness

thefull

capabilitiesof

trustworthy

artificial

intelligence

toadvance

United

States

diplomacy

andshape

thefuture

of

statecraft.E

A

I

SGOALS

&

OBJECTIVES1

2Leverage

Secure

AI

InfrastructureFoster

a

Culture

thatEmbracesAITechnologyIntegrateAItechnologiesintoasustainableandsecure

AI-enabling

infrastructureto

buildandscaleavarietyofAIapplications

acrosstheDepartment.Empower

adynamic

workforcewhosediverseneeds

for

AIareservedthroughtraining,aculture

of

continuouslearning,andhiring

forin-demand

AIskillsin

waysthatuphold

thehighestlevelsof

dataandscientific

integrity.1.1EnableAITechnologyIntegration1.2FullyUtilize

Infrastructure

for

AIAdoption

atDepartmentScale2.1ProvideAITraining

andSupport

Services2.2Develop

NewOpportunitiesfor

AI

Talent2.3PromoteResponsible

AIUse1.3Modernize

AcquisitionofAITools3

4Ensure

AI

is

Applied

ResponsiblyInnovateEstablish

theenterprise

capacitytoensuretrustworthy

andethical

AIuse,manageIdentify,

experiment,

andscalearange

ofsuccessfulsolutionstobe

anactiveinnovatorinapplied

AI,

while

forming

creativepartnershipswith

responsible

innovatorsoutside

of

theDepartmenttocompoundoursuccesses.algorithmic

risk,andassessdata

qualitywhileprovidingappropriate

accessto

AI-readydatato

informdecision-makingand

operations.3.1Establish

andMaintainAIGovernanceandPolicy4.1IdentifyOpportunities3.2BrokerAppropriateAccesstoAI-Ready4.2Facilitate

ResponsibleExperimentation4.3ScaleSuccessesData3.3Facilitate

DataQuality

Assurance<<

RETURN

TO

TABLE

OF

CONTENTSU.S.DepartmentofStateEnterpriseAIStrategyPage5INTRODUCTIONThe

Department

ofState

stands

at

acriticaljuncture

where

anemergingecosystem

ofartificialintelligence(AI)

capabilitiespresents

enormousopportunity.projects

that

willexemplify

responsible

andsecureAIadoption

willenableustoshape

ourtechnological

landscape

whilemitigatingsecurity

risksandbiases.Strategicallyinvestinginourapplied

AIcapabilities,

supportinginstitutions,andourworkforcetomodernize

theway

weservetheAmerican

peopledirectlysupports

Goal

4oftheFY2022

–2026

DepartmentofStateand

U.S.AgencyforInternational

Development(USAID)Joint

Strategic

Plan

(JSP),Part2ofthe2022National

Security

Strategy(NSS),andNationalSecurityMemorandum3(NSM-3)focused

onmodernizing

thenational

security

workforce,institutions,

andpartnerships.

Itisourdutytokeeppace

with

theglobalevolution

oftechnological

advancements.

TheDepartment’ssuccess

reliesonourability

toextractThisopportunity

can

allow

theDepartmenttoleverageAItoachievebreakthroughs

ofallkinds–inpublicdiplomacy,languagetranslation,managementoperations,information

proliferation

anddissemination,task

automation,codegeneration,andothers.Everyorganization

intheDepartmentcanbenefitfromAI,andsome

alreadyhave.

Manycan

usegenerativeAIandotherevolving

AItoolstoreap

immediaterewards

ofgreaterproductivity.

Increased

automation

ofroutinetaskswill

make

many

ofourdaily

efforts

moreefficient,

reducingworkload

burdeninWashington

andatour200+

postsworldwide,grantingmoretimeandfocus

onmission-critical

work.

Itisimperativethat

theDepartment

activelyleverage

AItoolsincarryingoutitsmission.

Wieldedintherightway,andwith

adequate

safeguards,

AIsystemscan

notonlyacceleratethe

pace

ofourinformation

gathering,

foreign

policy

analysis,andadvocacyon

behalf

oftheAmerican

people,butalso

helpusbemorediscerninginthosedomainsand

moreattentive

tothecriticallyhumanelements

ofourtrade.meaningful

insights

from

ourdata

tomakeinformeddecisionsand

supportmissiondelivery.ThelaunchofitsEnterprise

DataStrategy

(EDS)inSeptember2021

accelerated

theDepartment’s

appliedAIjourney.TheEDS

laidthefoundation

fortheEAIS

byprioritizinginvestment

intosecure,technical

infrastructureanddatagovernanceprocesses,Department-wideaccess

toaccurate

data,andtheempowermentofour

workforce

throughdata-centrictrainings.

Throughimplementation

oftheEDS,theDepartmenthasalready

madestridesindeployingAIapplicationsthat

haveresponsibly

freedourworkforce

from

routinerepetitive

laboranddeliverednewinsightstodecision-makers.Thisopportunity

willrequiretheDepartmenttotake

steps

to

protectindividuals

andmanagerisks,includingthesecurity

andprivacy

ofDepartment

data,whileworkingtoavertbiasedoutcomes

thatpose

arisktoourmission

andourvalues.

Thefirststep

fortheDepartmentinAIriskmitigation

willbeproactivelyprioritizingsecurity,bringingawarenesstothe

risksposedbyAI,andestablishing

clearDepartment

widegovernance

andpoliciesfor

responsible

AI.Fromthere,providingAItrainingandselectivelyinitiatingAIThisfirst,two-yearEnterpriseAIStrategy(EAIS)buildsoff

that

success

and

willreach

forthestillgreateropportunitiesAI

systemsofferwhenusedresponsibly.

Itis

theproductoftheStateDepartment’s

AIleaders

andpolicy

expertsfromover25

bureausandofficesacrosstheenterprise,dedicated

toensuringthat

ouradoption

ofAIisworthyoftheDepartment’sexcellence

indiplomacyand

foreignpolicy.<<

RETURN

TO

TABLE

OF

CONTENTSU.S.DepartmentofStateEnterpriseAIStrategyPage6STRATEGIC

DRIVERSAI

is

transforming

the

state

ofglobalaffairs,

and

the

conductofU.S.foreign

policymust

evolve

tomaintainleadership

onthe

worldstage.globe,inservice

tothecountry.

In

thecomingyears,ourabilitytogenerate,

translate,

andsummarize

language,improve

access

toinstitutional

knowledge,understand

themotivationsofourcompetitors,process

everincreasing

andoverwhelmingvolumesofinformation,

andcoordinate

theactivities

ofroughly270

diplomaticposts

inabout

180countries

willhave

directbearingonthestatureoftheUnited

Statesand

thewellbeing

ofAmericans.Thedevelopmentofthe

Department’s

firstEAISreflects

aworldtakingnewshape

throughtechnological

innovation

and,likewise,

anagency

fullofopportunity

inresponsible

AI.GlobalProliferationof

TechnologicalInnovationToday,

weseeimmense

opportunity

inappliedAItosupportthepeople

carryingoutthatmission.

Indomainswhere

Department

teamshavealready

proventhevalue

ofresponsibleAI,wecan

refineourabilitiesand

scale

upouraspirations.

In

parallel,weshouldresponsiblypressintonewdomains,lookingtoleverageemergingAIapproaches

which

can

servethepeopleservingtheUnitedStates

intheirmanycapacities,whetherseated

atnegotiating

tables,maintainingembassyfacilities,

orsecuringournetworks.ThedisruptivepowerofAIhasasserted

itselfacrossmanyfacetsofgovernment,

business,andeveryday

life,thanks

inlargeparttotheproliferation

andavailabilityofuser-friendlygenerativeAI.

Usersnolongerneedvastresources

ortechnicalexpertise

toleverageAI.Today’s

rapidcreation

ofneweasy-to-useAIsystemswill

permanentlyalter

thewayorganizationsoperate

andleadintheinformation

economy.Paramount

tounlocking

those

successes,

anditself

acenterpiece

ofourstrategy,

isouracutesenseoftherisksembeddedinthisdomainandourdutytounderwriteeveryusecase

withtheprinciplesofsafe,

secure,trustworthyAI,ensuringAItechniques

areusedinaccurateandappropriatemanners,

whileupholdingintegrity.AIsystemsarepoised

totake

onanewandprofound

prominence

inglobal

affairs,

andtheDepartment

must

beat

theforefront.

Just

asthebirthoftheinternetbentthepathofdiplomacytoward

adifferent

future,emergentAItechnologieswill

color

theworldinwayswehadnotimagined

justafewyears

ago.Todevelopandimplement

U.S.

foreignpolicyinthat

worldnewlyshaped

byAI,theDepartment

itselfmustthoughtfullygrow

itsAIcapabilities

andtraining,understandandmitigateitspractical

risks,andconnect

ourpractice

toourpolicy.Asthefoundation

oftheDepartment’s

AIreadiness,theEDS

hasprovideda

frameworktoleveragedata

byempowering

ourworkforce,providinggreateraccess

toaccurate

data,

andinvestingintheDepartment’s

technicalinfrastructureanddata

governance

processes.TheEDS

hasprovidedusaninstitutionalfoothold

toaccelerate

AIadoption,andtheingenuityofourpersonnelinpursuitofourcollective

mission

issimultaneouslydrivingenterprise-widedemand

forAI.Thisstrategyseeks

toservethat

demand

bytakingadvantage

ofthisonce-in-a-generationopportunity.Opportunities

for

ResponsibleAITheDepartmentis

chargedwithconductingU.S.

foreignpolicy,

atradeinwhich

informationiscurrency,andwhose

coordination

involvesnearly80,000

personneldeployed

aroundthe<<

RETURN

TO

TABLE

OF

CONTENTSU.S.DepartmentofStateEnterpriseAIStrategyPage7DEPARTMENT

OF

STAT

E

AI

VISIONVISION

STATEMENTTheDepartmentofState

willresponsibly

andsecurelyharness

thefullcapabilitiesoftrustworthyartificial

intelligence

toadvance

United

Statesdiplomacyand

shapethefutureofstatecraft.FROM

VISION

TO

ACTIONToguidetheDepartmenttowardsitsVision,fourGoalsserveasfoundationaltargets

thatwillenhance

theDepartment’s

AIcapabilities.Each

Goal

restsonspecificobjectives

thatencompassprioritiesidentified

bytheDepartment’s

AIleaders.

Theserelevant

andachievableefforts

willenablemeasurable

advancementoverthenexttwo

years.“More

than

anything

else,

our

task

is

to

put

forth

and

carryout

a

compelling

vision

for

how

to

use

technology

in

a

waythat

serves

our

people,

protects

our

interests,

and

upholdsour

democratic

values.

It’s

not

enough

to

highlight

thehorrors

of

techno-authoritarianism...

We’ve

also

got

tomake

the

positive

case

for

our

own

approach,

and

thenwe’ve

got

to

deliver.”Thatis

the

challengebefore

us.-

S

E

C

R

E

T

AR

Y

AN

T

O

N

Y

J

.

BL

I

N

K

E

N<<

RETURN

TO

TABLE

OF

CONTENTSU.S.DepartmentofStateEnterpriseAIStrategyPage8GOAL

1L

E

V

E

R

A

G

E

S

E

C

U

R

EA

I

I

N

F

R

A

S

T

R

U

C

T

U

R

EIMPERATIVESOLUTIONIMPACTTheDepartment

of

Stateworkforce

requiresbest-in-class

software

capabilitiesthat

leverageAI

andensure

security,

scalability,andease

of

use.Integrate

AI

technologiesintoa

sustainable

andsecure

AI-enablinginfrastructure

to

buildandscale

avarietyof

AIapplicationsacross

theDepartment.Providingsecure,

safe

technologyenvironments

to

developersandtransparent

tools

to

standard

userspromotes

equitableAI

use,

ensuresdata

protection

andprivacy,safeguards

against

potentialcybersecuritythreats,

andprotectsindividuals.Objective1.1:

EnableAITechnologyIntegrationindataencryption

mechanisms,robustnetworksecurity,multi-factorauthentication,

andregulardata

backupstosafeguarditsdata.Tobuildandscale

avariety

ofAItechnologies,theDepartment

willintegrateimpactful

AItechnologiesintosustainable,AI-enablinginfrastructure,withsecurity

asatoppriority.ItwillseektoprovidebroadaccesstoAItechnologies,commensurate

withtheDepartment’s

rangeofuserabilities,

with

amixofopen-source,

commercially

available,

andcustom-builtAI

systems.Robustaccess

controlsandauthentication

mechanismsaligned

toZeroTrustprincipleswillmitigate

riskofObjective1.3:

ModernizeAcquisitionof

AIToolsTheDepartment’s

IT

procurementauthorities,inpartnershipwiththeChief

Dataand

AIOfficer(CDAO),ResponsibleAI

Officer(RAIO),ChiefInnovation

Officer(CIO),andothers,willworktostreamline

theapproval

andprocurementofprioritizedAItechnologiestomeet

thedemandsignaledbytheDepartment’s

strongestunauthorized

access

toAItechnologies

andDepartment

data,providingahighlevelofsecurity.potential

usecases,

consistentwith

applicablelaw

andregulation.

ThisincludesexaminingtheIT

procurementpipelinetofindefficiencieswhilemaintainingthesafeguards

providedintheFederal

Risk

andAuthorization

ManagementProgram(FedRAMP),theInformationObjective1.2:

FullyUtilizeInfrastructureforAIAdoption

atDepartmentScaleTechnology

ChangeControl

Board(ITCCB),Authorization

toOperate

(ATO)process,

andotherapproval

mechanisms,

following

federalguidelines.

Toprioritize

investmentopportunities,

theDepartmentwill

identify

usecaseswhere

AIcan

providethehighestimpact,utilizingcoordination

bykey

offices

liketheCDAO

andissued

FedRAMP

frameworks.

Priortotheiracquisition,

AItechnologies

willbeevaluated

againstsecurityprotocolsand

riskassessmentprocesses.

TheDepartment’sprocurementandopen-source

approvalprocesseswill

befurtherdeveloped

toenableflexibility

andstreamlined

deploymentofAI.Evaluatingvendorclaimsand

developingnewlanguageforDepartment

contractswill

ensureourpartnersareheldtothesame

standards

ofsecurity,riskmanagement,

andtransparency

toenshrinetheserequirementsgoingforward.TheDepartmentwill

relyonarobusttechnologyinfrastructurethat

furtherenables

computing,development,

testing,deployment,andcontinuous

monitoringofAItechnologies,

whilealso

protecting

Department

data

andsecurity.Leveragingresources

withintheBureauofInformation

ResourceManagement(IRM)andintegration

with

theInformation

TechnologyExecutiveCouncil

(ITEC),theDepartmentwilldesignandimplementsupplementarytechnology

architecture

that

allows

forintegration

ofAIcomponentsintoourexistinginfrastructureanddata

pipelines.

Tomeet

thecomputational

demands

ofAIdevelopment,ourinfrastructurewillleverageDepartmentcloud-based

solutions

andscalable

infrastructureservices.

TheDepartmentwillrely

onexpertise<<

RETURN

TO

TABLE

OF

CONTENTSU.S.DepartmentofStateEnterpriseAIStrategyPage9GOAL

2F

O

S

T

E

R

A

C

U

L

T

U

R

E

T

H

ATE

M

B

R

A

C

E

S

A

I

T

E

C

H

N

O

L

O

G

YIMPERATIVETheDepartment’sSOLUTIONIMPACTEmpower

adynamic

workforcewhosediverse

needsfor

AIareserved

through

training,IncreasedAIfluencyworkforceneedstheskillset,

robusttrainingcurriculum,

enhanced

AIliteracy,

andeducationalresourcesto

ensureAIistobewidelyand

safelyadoptedacross

theDepartment.enables

allemployeestosafelyand

responsiblyreduce

time

spent

ontasksoptimal

for

AIpromotingacultureofcontinuouslearning,andhiringfor

in-demand

AIskillsin

waysthatuphold

thehighestlevelsof

dataandscientific

integrity.solutions,allowingthemto

focustheir

efforts

onhigher-impact

activities.Objective2.1:

ProvideAITrainingandSupport

Servicesprinciples,andethics,whocan

playpivotal

rolesinouradoption

ofresponsible

AI.

TheDepartment

will

buildonitssuccesshiringacohort

ofdata

sciencepractitioners

under

theguidanceoftheDepartment’s

CDAO,andopenfreshopportunitiesfor

technical

practitionersthroughthedevelopment

ofnewAIfocusedroles,such

asdatascientists,

operationsresearchers,andIT

specialists,

anddeploymentofprogramstosupport,attract,andretainAItalent.TheDepartmentwill

develop

specialized

AIlearningopportunitiestomeetthe

diverseneedsofitsworkforce,

enhanceAIliteracy,encourageandeducate

onresponsible

AIuse,andensurethat

userscan

adequatelymitigatesome

risksassociated

with

AItools.

AsAIisintegratedintoDepartmentinfrastructureandexistingtechnology

platforms,

itisintegralthatourworkforce

understands

what

thesetechnologiesare

andhowtosafely

useandapplythem.

TheDepartmentwillincrease

AIfluency

forbothtechnical

andnon-technicalusersthroughthedevelopmentofmulti-tiered,incentivized

trainingsandmodificationsofexistingtrainings,ledbytheForeignServiceInstitute

(FSI).Tofurthersupportusers,technology-specificmaterialswill

bedevelopedtoassistintherecognition,exploration,andinterpretation

ofAI,as

well

asthefacilitation

ofsupportsessionsto

assistallAI

users.

TheDepartment

will

convene

communities

ofpractice

toshareAIresources,usecasesandbestpractices,

anddevelopment

ofspecificimpactmetricstoaccompanyAI

technologiesthat

willestablish

parameters

fortheexpectedbenefits

ofuse.Objective2.3:

Promote

ResponsibleAIUseInthisearlystage

ofAIadoption,

theDepartment

must

wrestle

with

extensiveunknowns

tonavigate

thepathofopportunitywhileensuringresponsible

AIpractice,includingbyrespecting

andpromoting

security,privacy,equity,andothercoreprinciples.

Much

liketheEDS

aims

tocultivatea

dataculture,theDepartment

will

imbueitsvalues

aroundresponsible

AIuseacrossthe

organization,includingtoupholddata

andscientificintegrity.TheDepartmentwill

instill

bestpractices

thatroutinizes

responsible

AIusebyteaching

ourworkforcewhen

andhowtouseAItoolseffectively,safely,

andlawfully.

Throughthedevelopmentofinterdisciplinary

trainingcourses,wewilleducate

ourworkforce

onthebasicsofAI

riskandriskmitigation

techniquestoempowereffective

AIuseandupholddataandscientific

integrity,whilealso

recognizingthelevel

ofacceptable

riskthat

accompanieseach

AIapplication.

Wewill

comply

withapplicable

lawandAIgovernance

andpolicyguidelinesandminimizetheriskofAIuse.Objective2.2:

DevelopNewOpportunitiesfor

AITalentTheDepartmentwill

recruitandhireAIexpertisecurrentlyunderrepresentedinitsworkforce,especially

thosewithanunderstandingofAItechniques,technologies,<<

RETURN

TO

TABLE

OF

CONTENTSU.S.DepartmentofStateEnterpriseAIStrategyPage

10GOAL

3E

N

S

U

R

E

A

I

I

S

A

P

P

L

I

E

DR

E

S

P

O

N

S

I

B

L

YIMPERATIVESOLUTIONEstablish

theenterprise

capacitytoensuretrustworthy

andethicalAI

use;ensuredata

privacyandsecurity,manage

algorithmicrisk;

reduceharm,bias,

andinequity;andassess

dataqualitywhileproviding

appropriateaccess

to

AI-ready

data

to

informdecision-making

andoperations.IMPACTTheDepartment

musthavepolicy,guidelines,andtechnologies

tofacilitate

transparent,responsible,

andethicaluse

of

applied

AI,Responsible

data

managementandAI

governance

ensures

theavailabilityof

high-quality,trustworthy,

andrelevantdata,creatingafoundation

foraccurate

model

development,reliableinsightgeneration,

andresponsibleandethicaluse.includingcontinualevaluationandtesting.Objective3.1:

Establishand

MaintainAIGovernanceandPolicythreats,mitigate

bias,andensuredataprotection.

Thepoliciesand

guidelinesdeveloped

willconsider

thesecurity

andprivacyofDepartment

data.

TheDepartment

will

ensureclearandtransparent

proceduresforlegalandpolicy

reviewofnewAIusecases.TheDepartmentwill

oversee

andmanagerisk,adheretotheprinciples,guidelines,tools

andpracticesestablished

fromkey

directives

(e.g.,ExecutiveOrders),anddevelop

additional

policytoensurealignment

ofappliedAIwithapplicablelaw

andpolicy,

andwith

standards

forObjective3.2:

BrokerAppropriateAccesstoAI-readyDataresponsible

andethical

use,throughtheTheDepartmentwill

streamline

andensureappropriate

access

tointernal,interagency,andthird-partydata

for

AIusethat

istransparentlysourced.

TheData.Stateplatformwillprovideenterprise-wideaccesstodatawhen

possibleandappropriate,consistent

withapplicable

lawandprotections.

Safeguards,protocols,

anddatamanagementstandardsapply

wherenecessary,inaddition

todata

sharingagreements

thatreflect

theDepartment’s

policiesfordata

useinitstechnology

platformsand

with

vendors.EnterpriseData&

AICouncil

(EDAC),theAISteeringCommittee

(AISC),andtheDataGovernanceNetwork.

TheDepartment’s

CDAOwillsupportandcoordinatethe

establishmentandmaintenance

ofAIpolicies—such

as

20

FAM201.1—thatprovideclearguidelinesforresponsible

AIuse,stewardAImodels,

andprioritizetheevaluation

andmanagementofalgorithmicrisk

(e.g.,

risksarisingfromusingalgorithms)inAIapplications

duringtheirentirelifecycle—

includingthose

relatedtorecordsretention,privacy,cybersecurity,

andsafety.Thatcommitmentimplicatesmanydatasciencedisciplines,likedata

collection,

extraction,transformation,

andloading;

model

selection,development,

deployment,

andmonitoringinproduction;

statistical

methods;

andothers.

AIcomplianceplansand

protocolsforsystemmaintenance,recalibration,andusestoppagewillpreventunintendedbias

andfunctionality.Minimum

risk-managementpractices

forrights-andsafety-impactingAIwillbeestablished

fordevelopment

and

procurement.TheRAIO,perCDAOdirection,