Copy igh © ISRG Publishe s. All igh s Rese ed.
DOI: 10.5281/zenodo.17283672
146
ISRG PUBLISHERS
Abb e ia ed Key Ti le: Is g J Econ Bus Manag
ISSN: 2584-0916 (Online)
Jou nal homepage: h ps://is gpublishe s.com/is gjebm/
Volume – III Issue -V (Sep embe -Oc obe ) 2025
F equency: Bimon hly
The impac o a i icial in elligence on public adminis a ion in Ame ica
DR. Khai eya Redwan Yahya
Al-Is iqlal Uni e si y, Je icho, Pales ine
| Recei ed: 23.09.2025 | Accep ed: 29.09.2025 | Published: 07.10.2025
*Co esponding au ho : DR. Khai eya Redwan Yahya
Al-Is iqlal Uni e si y, Je icho, Pales ine
INTRODUCTION
F om policy-making o se ice deli e y, a i icial in elligence (AI)
is oday an in eg al pa o he con empo a y go e nmen , a ec ing
mos aspec s o public adminis a ion. In an a emp o s eamline
he sys em, enhance decision-making, and op imize he use o
esou ces, Uni ed S a es go e nmen agencies inc easingly ely on
a i icial in elligence echnology. On he o he hand, i s apid
in eg a ion has aised genuine ques ions in e ms o e hical issues,
anspa ency, and subs i u ion o human labo (Nade & e al,
2024). Figu e 1 shows A Comp ehensi e Re iew o A i icial
In elligence’s Impac on Decision-Making acco ding o (Caiza & e
al, 2024).
Th ough c i ical examina ion o i s ad an ages, disad an ages, and
u u e es ima ed implica ions, he objec i e o his pape is o shed
ligh on he a ious implica ions a i icial in elligence has
in oduced in o public adminis a ion (Wang & e al, 2024).
Despi e he ac ha he applica ion o a i icial in elligence in
public adminis a ion is no a new phenomenon, he accep ance o
his echnology has accele a ed o e he cou se o he p e ious 10
Abs ac
The ma e nal ma ine supplemen s sec o has ga ne ed signi ican a en ion in ecen yea s due o g owing awa eness o he c i ical
impo ance o ma e nal heal h and nu i ional suppo . This s udy aims o examine he implemen a ion o G een Supply Chain
Managemen (GSCM) p ac ices wi hin his indus y as an app oach o achie ing sus ainable global comme ce. The esea ch employs
a quali a i e app oach, conduc ing comp ehensi e e iews o cu en li e a u e and examining case s udies o iden i y e ec i e
s a egies and obs acles ela ed o GSCM adop ion in he ma e nal ma ine supplemen s ield. Resul s demons a e ha implemen ing
GSCM p ac ices yields mul iple ad an ages, including enhanced en i onmen al p o ec ion, s onge inancial pe o mance, and
inc eased consume con idence. The s udy concludes by o e ing s a egic guidance o inco po a ing GSCM amewo ks in o he
ope a ional p ocesses o supply chain pa icipan s.
KEYWORDS:
G een Supply Chain Managemen , Ma e nal Ma ine Supplemen s, Sus ainable T ade, En i onmen al
Sus ainabili y, In e na ional T ade
Copy igh © ISRG Publishe s. All igh s Rese ed.
DOI: 10.5281/zenodo.17283672
147
yea s (Yigi canla & e al, 2024). Au oma ion o adminis a i e
du ies is one o he many applica ions o a i icial in elligence.
O he applica ions include policy insigh analysis o massi e
amoun s o da a (Mohammed & e al, 2024). Ini ia i es d i en by
a i icial in elligence a e ecei ing signi ican inancing om bo h
he ede al go e nmen and s a e go e nmen s ac oss he Uni ed
S a es (He mann & e al, 2024). These ini ia i es include p edic i e
analy ics o law en o cemen , machine lea ning algo i hms o
de ec ing aud in wel a e p og ams, and cha bo -based se ices o
public in e ac ion. Fo all hese de elopmen s, conce ns abou
algo i hmic bias, da a p i acy, and he absence o human judgmen
s ill linge . Acco ding o indings in p e ious s udies, he e a e
ins ances when a i icial in elligence sys ems unin en ionally
ein o ce sys ema ic biases o make mis akes because hey a e
ained on incomple e da a se s (Adewusi & e al, 2024).
Fu he mo e, e en while a i icial in elligence p esen s he p omise
o g ea e e iciency, he use o his echnology p esen s some
go e nance and e hical challenges ha will need o be ca e ully
esol ed (Fosso Wamba & e al, 2024). Policymake s need o each
an unde s anding o he complexi y o he echnologies hemsel es
i s , p io o being able o an icipa e using a i icial in elligence in
public adminis a ion in an app op ia e manne (Schwei ze , 2024).
Figu e 1: A Comp ehensi e Re iew o A i icial In elligence’s
Impac on Decision-Making acco ding o (Caiza & e al, 2024)
Al hough he applica ion o a i icial in elligence p omises a weal h
o oppo uni ies o imp o ing public adminis a ion, he
deploymen o his echnology is augh wi h challenges ha
equi e ca e ul conside a ion. Th oughou he cou se o his e o ,
he ollowing esea ch ques ions a e add essed:
- In wha ways is a i icial in elligence (AI) cu en ly
being in eg a ed in o a ious public adminis a ion a eas
ac oss he Uni ed S a es o Ame ica?
- In wha ways could a i icial in elligence inc ease he
e iciency o he go e nmen , pa icula ly wi h ega d o
decision-making and he deli e y o se ices?
- When i comes o adminis a i e sys ems d i en by
a i icial in elligence, wha ac o s in luence public
con idence, and how do people pe cei e such sys ems?
Rela ed Wo ks
A i icial in elligence has enhanced he managemen o da a in he
public sec o , he de ec ion o aud, and p edic i e analysis. To
comp ehend such in icacies, an app ecia ion o he impac
a i icial in elligence has on he e iciency o adminis a ion,
implemen a ion o policy, and decision-making is necessa y. Wi h
ega ds o me hods, esul s, and a eas o addi ional esea ch, his
sec ion examines some o he mos signi ican AI esea ch o
public adminis a ion. The pu pose o his e iew is o pu he
applica ion o AI in go e nmen in o pe spec i e by compa ing
p e ious and mo e ecen empi ical indings in he hopes o
de e mining a eas o lacking esea ch.
Maybe mo e han e e be o e, e e yday li e, socie y, and s a e
sys ems a e impac ed by AI en i onmen s and algo i hms.
Go e nmen s ha e la gely been esponsible o making su e ha
oday's public adminis a ion sys ems a e in place and acili a ing
smoo h ansi ioning in o new echnologies, as p esen ed by Uzun
(2022). In o de o in eg a e, go e n, and egula e AI echnology,
public adminis a ion and policy ha e impo an oles o play.
The e a e se e al oppo uni ies when AI is included in o public
adminis a ion and he policy-making p ocess. In his sense, he
big ques ions deba e began in 1995 when Robe Behn highligh ed
he impo ance o he big ques ions as he main mo i a o o a
esea ch p og am in public adminis a ion. As a esul , while AI
can assis in sol ing kno y global challenges, i gene a es se e e
p i acy, accoun abili y, and egula o y issues. This s udy
demons a es he poli ical, legal, and public adminis a ion aspec s
o ansdisciplina y AI egula ion. Policymake s need o add ess
AI elen lessly o maximize i s ad an ages while minimizing i s
disad an ages (Uzun & e al, 2022).
Al hough he use o a i icial in elligence (AI) cha bo s in public
o ganiza ions has inc eased in ecen yea s, h ee c ucial gaps
emain un esol ed. Fi s , li le empi ical e idence has been
p oduced o examine he deploymen o cha bo s in go e nmen
con ex s. Second, exis ing esea ch does no dis inguish clea ly
be ween he d i e s o adop ion and he de e minan s o success
and, he e o e, be ween he s ages o adop ion and implemen a ion.
Thi d, mos cu en esea ch does no use a mul idimensional
pe spec i e o unde s and he adop ion and implemen a ion o AI
in go e nmen o ganiza ions. o his ega d, Chen 2024 add essed
hese gaps by explo ing he ollowing ques ion: wha de e minan s
acili a e o impede he adop ion and implemen a ion o cha bo s in
he public sec o ? his ques ion by analyzing 22 s a e agencies we e
answe ed ac oss he U.S.A. ha use cha bo s. I showed ha
di e en ypes o de e minan s (such as knowledge-based c ea ion
and main enance, echnology skills and sys em c ashes, human and
inancial esou ces, c oss-agency in e ac ion and communica ion,
con iden iali y and sa e y ules and egula ions, and ci izens’
expec a ions, and he COVID-19 c isis) impac di e en ly he
adop ion and implemen a ion p ocesses and, he e o e, de e mine
he success o cha bo s in a di e en manne . Table 1 shows he
De e minan s o AI Cha bo Implemen a ion Success as p esen ed
(Chen & e al, 2024).
Table 1: De e minan s o AI Cha bo Implemen a ion Success acco ding o (Chen & e al, 2024)
Fac o
Second-o de Code
Challenge (Fi s -o de Code)
Enable (Fi s -o de Code)
Da a &
Knowledge-based
Limi ed websi e analy ics use (10)
Collabo a e wi h se ice s a (3)
Copy igh © ISRG Publishe s. All igh s Rese ed.
DOI: 10.5281/zenodo.17283672
148
In o ma ion
c ea ion
Con en enhancemen
Use s ph ase ques ions di e en ly (4)
Moni o cha bo pe o mance (3, 2)
Technology
Technology skills
Lack o cha bo de elopmen skills (12)
Pa ne wi h IT endo s (12)
Sys em
Lack o cha bo managemen skills (4)
P o ide aining (2)
Sys em c ashes (2)
Use a so launch app oach (2)
O ganiza ional
Human esou ces
Inc eased wo kload (9)
Realloca e s a (3)
Financial esou ces
Long- e m unding sus ainabili y (2)
Use ee ials, con ac s, o special unding
(12)
In e -
o ganiza ional
S a expec a ions
Deploymen seen as complex & bu densome (2)
Imp o e c oss-agency communica ion (2)
Execu i e expec a ions
Un ealis ic expec a ions abou cha bo s (2)
Cla i y p ojec scope (2)
Ins i u ional
Con iden iali y & sa e y
No PII equi ed o cha bo use (11)
Con ex ual
End-use expec a ions
Use s expec speci ic de ails (2)
Communica ion & cla i ica ion (2)
Examining how AI in e ac s wi h public go e nance and
managemen in bo h de eloped and eme ging ma ke economies
was he goal o ano he s udy. The e o e, he pape made he case
ha a i icial in elligence (AI) has eno mous po en ial and has been
used o imp o e go e nmen pe o mance in he ollowing a eas:
poli ics, in e go e nmen al ela ions, policymaking, social se ice
deli e y, public secu i y managemen , public inancial
managemen , in o ma ion p ocessing, and da a managemen . The
s udy came o he conclusion ha because AI o e s a ac i e
e u ns on in es men , he e a e s ill a lo o un apped po en ials in
he in e ac ion be ween AI and public adminis a ion. As a esul ,
mo e esea ch should be done and p o essionals who a e in e es ed
in in eg a ing AI in o mo e unc ional a eas o public managemen
and go e nance should ecei e mo e suppo (Agba & e al, 2023).
Using a da ase o 3,149 documen s om he Scopus da abase,
ano he s udy iden i ied he op 200 mos ci ed a icles based on
annual ci a ions. In addi ion, selec ed AI use cases om he
Eu opean Commission da abase we e classi ied, ocusing on hei
con ibu ions o public alue. The analysis ocused on h ee
dimensions o go e nance: in e nal p ocesses, se ice deli e y, and
policy making. The esul s p o ided a ca ego ized unde s anding o
AI concep s, ypes, and applica ions in he Pales inian Au ho i y,
along wi h a discussion o bes p ac ices and challenges (Babšek &
e al, 2025).
The connec ion be ween disc e ion, bu eauc a ic o m, and
a i icial in elligence (AI) in public o ganiza ions was he subjec
o ano he essay. The s udy came o he conclusion ha while he
usage o AI has g own ecen ly, li le is known abou how his
unique kind o ICT, disc e ion, and bu eauc a ic o m ela e o one
ano he along he s ee -le el o sys em-le el con inuum. As a
unc ion o bu eauc a ic o m, he s udy disco e ed ha he impac
o AI on disc e ion is nonlinea and nonmono onic. Howe e , e en
i hese o ganiza ions ha e p e iously opposed such changes, he
deploymen o AI may accele a e hei shi om s ee -le el and
sc een-le el bu eauc a s o sys em-le el bu eauc acies (Bullock &
e al, 2020).
The goal o Schi , 2020, was o e alua e he co e public alues
ha could be jeopa dized by he adop ion o au oma ed decision-
making sys ems (ADS) by go e nmen s. In o de o quan i y and
compa e he signi icance o h ee o hese alues— ai ness,
anspa ency, and human esponsi eness—a public alue ailu e
amewo k and empi ical app oach we e used. The esul s o a
su ey expe imen conduc ed among ields o c iminal jus ice and
child wel a e demons a ed unequi ocally ha some public alue
ailu es ela ed o AI ha e a subs an ial de imen al impac on
ci izens' assessmen s o hei go e nmen . They disco e ed ha
when ai ness and openness a e no me in he applica ion o ADS,
he e a e no able un a o able eac ions om he public. E en in
cases whe e esponden s we e no di ec ly impac ed, hese esul s
held ue ega dless o poli ical backg ound o ideology (Schi &
e al, 2022).
In spi e o he ex ensi e esea ch on a i icial in elligence in public
adminis a ion and go e nance, he e a e s ill gaps in esea ch such
as he lack o esea ch on anspa ency, e hics, and educing bias in
AI applica ions and a lack o compa a i e s udies ha e lec he
challenges o implemen ing AI in de eloping economies. In
addi ion, sho a e longi udinal measu es gauging he long- e m
e ec s o a i icial in elligence on go e nmen us and he public
sec o . Also lacking is much in es iga ion o he in e sec ion o
a i icial in elligence, decision-making, and accoun abili y in
public policy. Fu he mo e, public adminis a ion gene a i e AI
solu ions (like cha bo s) need imp o ed unde s anding o
implemen a ion p oblems as well as he pa icipa ion o
indi iduals, and hence addi ional applied and expe imen s udies
a e equi ed o ensu e e icien go e nance and clea e hics o AI
deploymen in he public sec o . Fo his ega d, he objec i es o
his esea ch s udy: o assess he le el and g ade o a i icial
in elligence (AI) up ake in public adminis a ion in he Uni ed
S a es; o explo e he ways in which AI impac s go e nmen al
e iciency, policymaking, and he dynamics o labo ; o iden i y he
p ime e hical, ju idical, and echnological ques ions ega ding AI
ollou ; and o p o ide policy ecommenda ions o how one can
ad ance he go e nance o AI. By means o a c i ical e alua ion o
he cu en a i icial in elligence deploymen in go e nmen al
ins i u ions, his s udy a emp s o examine he bene i s and
d awbacks o a i icial in elligence go e nance. Mo eo e , i
looks a how a i icial in elligence echnologies a ec
adminis a i e decision-making, esou ce alloca ion, and se ice
deli e y pe o mance. This pape in ends o highligh impo an
legal and go e nance ques ions esul ing om algo i hmic bias,
da a p i acy di icul ies, and human labou displacemen . The
p ojec seeks o o e e idence-based policy ecommenda ions o
gua an ee esponsible and e hical uses o a i icial in elligence in
public adminis a ion, so combining public con idence wi h
echnical de elopmen .
Copy igh © ISRG Publishe s. All igh s Rese ed.
DOI: 10.5281/zenodo.17283672
149
Me hodology
By examining i s e ec s on ins i u ional e iciency, decision-
making, anspa ency, and go e nmen in es men , his s udy seeks
o in es iga e how a i icial in elligence (AI) is a ec ing public
adminis a ion in he Uni ed S a es. In an e o o unde s and he
ad an ages and di icul ies o implemen ing AI in he public sec o ,
he s udy uses a mixed-me hods app oach, in eg a ing quan i a i e
and quali a i e da a (See Figu e 2). The s udy examines, using he
desc ip i e-analy ical app oach, he in luence o a i icial
in elligence (AI) on Ame ican public adminis a ion. The s udy
uses quali a i e and quan i a i e da a o examine he in luence o
a i icial in elligence on he pe o mance o ins i u ions, decision-
making, anspa ency, and go e nmen expendi u es. The s udy
employs h ee main sou ces o da a usage in an a emp o p o ide
accu acy and comp ehensi eness:
- O icial S a is ical Da a including DARPA's and he U.S.
Bu eau o S a is ics epo s on a i icial in elligence
in es men s and use (Na ional Academy o Public
Adminis a ion , 2019), (Cen e s o Medica e &
Medicaid Se ices, 2019), and (Ins i u e, 2019).
- Academic jou nals ensu e heo e ical unde pinnings
h ough Public Adminis a ion Re iew a icles,
In o ma ion Poli y, and he Jou nal o Public
Adminis a ion (Bullock & e al, 2020), (Schi & e al,
2022).
- Case s udies o a i icial in elligence applica ion in cou
sys ems, he heal hca e sec o , police, and o he
adminis a i e unc ions o e alua e success a es and
gaps.
Ma hema ical models a e used in an e o o ga he s a is ical
in o ma ion, clean and analyze i in his seconda y da a analysis
me hod. One o he mos impo an models used and p esen ed as
p esen ed in Equa ion (1):
Y_ =α+β_ +ϵ_ …………………………….…… (1)
ep esen he AI adop ion a e a ime, he end
coe icien , he e o e m, and he in e cep . I can iden i y whe he
AI adop ion in public adminis a ion ollows an inc easing,
dec easing, o s agnan end o e ime.
Se e al quan i a i e echniques a e used o examine ends in AI
adop ion:
- S a is ical Analysis: The analysis quan i ies he impac o
AI on go e nmen pe o mance using desc ip i e
s a is ics like mean (μ), s anda d de ia ion (σ), and
in e qua ile ange (IQR). In es men in AI g ow h is
in es iga ed using Equa ion (2), which ep esen s AI
in es men a ime :
………………. (2)
- T end Analysis: This analysis can be used in his s udy o
unde s and how he adop ion o a i icial in elligence is
e ol ing. The compound annual g ow h a e (CAGR)
can be calcula ed as in Equa ion (3), whe e he inal,
ini ial AI adop ion alue and numbe o yea s a e
, espec i ely:
…. (3)
- Compa a i e analysis: This me hod helps e alua e he
adop ion o AI ac oss di e en adminis a i e sec o s
such as law en o cemen , axa ion, and public heal hca e.
The e iciency a io (ER) o each o hese sec o s can be
e alua ed using Equa ion (4):
…… (4)
Figu e 2: Flowcha ep esen ing he me hodology o he s udy.
Resul s
This sec ion seeks o p o ide a c i ical assessmen o e idence on
he impac o AI on public adminis a ion based on quan i a i e
measu es. This sec ion syn hesizes he esul s o his s udy ac oss
ou a eas o impo ance: he use o AI by go e nmen agencies,
in es men ends, public us and anspa ency, and employmen
impac . The indings show ha 63% o go e nmen agencies
al eady use AI, and ano he 59% will implemen AI soon. This
e eals he inc easing accep ance o he ole o AI in imp o ing
adminis a i e e ec i eness. The implemen a ion a e o AI-based
decision-making is only 48%, sugges ing ha al hough AI is being
emb aced, i s in eg a ion in o decision-making p ocesses is s ill in
he making, as illus a ed in Table 2.
Table 2: AI Adop ion in Go e nmen Agencies
Me ic
Value
Uni
Pe cen age o go e nmen agencies using AI
63%
%
Pe cen age o agencies planning o adop AI
59%
%
AI-d i en decision-making implemen a ion a e
48%
%
An adop ion a e o o e 60% indica es ha AI has become a key
ool in public adminis a ion. Howe e , he 48% decision-making
implemen a ion a e means ha while AI is widely a ailable,
agencies may be hesi an o ely on AI o make c i ical go e nance
decisions, pe haps due o e hical conce ns, egula o y ba ie s, o
echnical limi a ions. The e exis s a gap be ween adop ion and
e ec i e use, signi ying he need o policies suppo ing he easy
embedding o AI in o decision-making.
While he US go e nmen spend on AI s ood a $6 billion in 2023,
i is an icipa ed o be $9 billion by 2025, g owing a a CAGR o
Copy igh © ISRG Publishe s. All igh s Rese ed.
DOI: 10.5281/zenodo.17283672
150
12%. This ype o in es men by he go e nmen indica es he
s a egic impo ance o he de elopmen o AI, which can be seen
om Table 3. The 12% CAGR also indica es he inc easing
signi icance o AI in go e nmen policy and p oposal submissions.
Fu he , he es ima ed cos o $9 billion by 2025 indica es how AI
is also being seen as a long- e m d i e o adminis a i e
e ec i eness, cos educ ion, and se ice au oma ion. The
inc eased expendi u e equi es a clea ly es ablished AI go e nance
amewo k o enable esponsible expendi u e, a oid duplica ing
e o s, and ensu e maximum e u ns on in es men s in AI.
Table 3: In es men in AI and Budge Alloca ion
Me ic
Value
Uni
Go e nmen AI in es men in 2023
$6B
Billion USD
P ojec ed AI in es men in 2025
$9B
Billion USD
Annual AI in es men g ow h a e
12%
%
Public us in AI decision-making is di ided, 48% o whom we e
us ing, wi h 38% o he public s ill suspicious o openness.
Fu he mo e, 27 AI go e nance b eakdowns we e epo ed, which
e lec enhanced mo e open go e nance a angemen s as seen in
Table 4. 48% us o e all means es ic ed AI decision-making
accep ance bu equi es g ea e openness and accoun abili y.
T anspa ency issues emain a signi ican conce n, as 38% o
complain s we e p ocess-based AI-d i en.
Table 4: Public T us and T anspa ency
Me ic
Value
Uni
Public us in AI-d i en decisions
48%
%
Complain s abou AI anspa ency issues
38%
%
AI- ela ed adminis a i e ailu es
27
Cases
The 27 epo ed ailu es o go e nance also e lec ha he use o
AI is no lawless and equi es cons an adjus men and moni o ing
o educe e o s and inc ease eliabili y. Imp o ing anspa ency
mechanisms, such as explainable AI models and open egula o y
policies, can po en ially es ablish mo e public us in AI-d i en
decision-making. While examining he impac o AI on jobs and
labo o ce, AI is signi ican ly changing he go e nmen
wo k o ce, whe e 50,000 employees ha e o be e ained in o de
o adap o AI-d i en changes. In addi ion, 30% o whi e-colla
jobs we e a ec ed, and 16 AI p ojec s di ec ly con ibu ed o job
displacemen , as shown in Table 5. I can be concluded ha
e aining 50,000 employees demons a es a p oac i e e o by he
go e nmen o add ess wo k o ce dis up ions caused by AI. While
he 30% o whi e-colla jobs a ec ed highligh he ans o ma i e
impac o AI on go e nmen employmen , wi h ou ine and
epe i i e asks being au oma ed. The 16 AI-led job displacemen s
sugges ha while AI c ea es e iciency, i also c ea es social and
economic conce ns ela ed o job secu i y. Go e nmen s mus
balance AI e iciency gains wi h job-p o ec ion policies, such as
e aining p og ams and AI-human collabo a ion models.
Table 5: AI Impac on Wo k o ce and Employmen
Me ic
Value
Uni
Numbe o go e nmen employees
e ained due o AI
50,000
Employees
Pe cen age o adminis a i e jobs
30%
%
a ec ed by AI
AI p ojec s leading o job displacemen
16
P ojec s
On he o he hand, he bene i s o AI inno a ion include be e
decision p ocesses bu he ad ancemen challenges leade ship in
adap a ion as well as c ea es jobs losses and e hical quanda ies.
This esea ch s udies possible p edic i e ou comes o AI e ec s on
public adminis a ion and deli e s s a egic ecommenda ions o
handling i s in ica e issues.
The as pace a which AI sys ems en e adminis a i e wo k has
spa ked wo ies abou how such in eg a ion a ec s wo k o ces and
hei p i acy as well as equal oppo uni ies among ci izens.
Au oma ion o public sec o jobs ha comple e epe i i e
ope a ions c ea es an unemploymen p oblem ha a ec s la ge
numbe s o wo ke s. AI-based go e nance equi es addi ional
conside a ion o e hical issues which a ec accoun abili y
measu es as well as c ea es doub s abou anspa ency while
dis o ing ope a ional decision sys ems. When AI su passes human
in elligence, i es ablishes unce ain social condi ions which makes
u u e ou comes unp edic able o socie y.
Public adminis a i e agencies ac oss ede al and s a e le els
emb ace AI deploymen s which esul in p ocess op imiza ion
oge he wi h imp o ed se ice deli e y and be e decision-cen ic
capabili ies. AI-powe ed cha bo s along wi h p edic i e analy ics
and au oma ed documen p ocessing sys ems ha e subs an ially
aised ope a ional speed. The implemen a ion o algo i hmic
go e nance aces ongoing di icul ies o p e en biases om
appea ing while main aining ai ness s anda ds (See Table 6)
acco ding o (Yigi canla & e al, 2021) (Gillespie & e al, 2023).
Table 6: Challenges o AI Applica ion in Go e nmen Agency
Go e nmen Agency
AI Applica ion
Challenges
IRS
F aud de ec ion &
ax compliance
Risk o alse
posi i es
DHS
Facial ecogni ion
o secu i y
P i acy
conce ns
DMV
Au oma ed
licensing & ehicle
egis a ion
Algo i hmic
bias isks
Go e nmen unding o AI esea ch and de elopmen con inues o
inc ease because AI has become i al o public adminis a ion
needs. Commissioned ede al unding along wi h ech pa ne ships
and AI-based in as uc u e de elopmen demons a es o icial
con i ma ion o pe manen AI implemen a ion. The ai
dis ibu ion o AI echnologies oge he wi h unding alloca ion
ma e s emain subjec o ongoing conce ns be ween s akeholde s
(Yigi canla & e al, 2021) (Gillespie & e al, 2023).
The public main ains us in AI decision sys ems when hey a e
anspa en o iew. The public has de eloped doub s ega ding
how AI pa icipa es in go e nmen al ope a ions because hey
wo y abou bo h un aceable p ocesses and iola ions o e hical
guidelines. The de elopmen o explainable AI combined wi h
public engagemen p og ams will build us o AI-d i en public
se ices, acco ding o he in o ma ion p esen ed in Table 7
(Yigi canla & e al, 2021) (Gillespie & e al, 2023).
Copy igh © ISRG Publishe s. All igh s Rese ed.
DOI: 10.5281/zenodo.17283672
151
Table 7: Pe cen age o Responden s (%) o Public Conce n
Public Conce n
Pe cen age o
Responden s (%)
AI bias in go e nmen decision-making
68%
Lack o anspa ency in AI policies
74%
P i acy isks in AI-powe ed se ices
81%
The in oduc ion o AI au oma ion echnologies has modi ied
public sec o labo o ces which led o pe sonnel layo in ce ain
posi ions bu in oduced posi ions dedica ed o AI managemen and
o e sigh esponsibili ies. The main issue es s in deli e ing p ope
aining p og ams o employees who los hei jobs o ensu e hey
can main ain ele ance in he AI-based economic en i onmen
ollowing da a in Table 8.
Table 8: Jobs C ea ed and Los Due o AI in Cus ome Se ice,
Da a En y, and Cybe secu i y
Sec o
Jobs Los Due o AI
Jobs C ea ed by AI
Cus ome
Se ice
120,000
40,000
Da a En y
90,000
30,000
Cybe secu i y
20,000
60,000
The ollowing h ee scena ios depic AI's pa h in public
adminis a ion oge he wi h hei e ec s on human occupa ion
acco ding o (Yigi canla & e al, 2021) (Gillespie & e al, 2023):
Scena io 1: Human Con ol O e AI and Robo ics
The u iliza ion o AI ools unde human o e sigh in his si ua ion
pe mi s public adminis a ion o bene i om echnological
ad ances wi hou losing human employmen posi ions. Public
adminis a ions enac con olled legisla ion o lead AI de elopmen
ac i i ies by moni o ing e hical p ac ices and main aining human
supe ision h oughou he p ocess. This me hod allows bo h
human con ol and AI e iciency o wo k oge he so ha people
main ain hei decision-making au ho i y. This scena io shows a
majo bene i because AI suppo s human wo ke s ins ead o
aking hei posi ions hus diminishing economic as well as e hical
complica ions.
Scena io 2: Coexis ence Be ween Humans and AI
Wi hin his model, AI o ms a mu ually bene icial pa ne ship wi h
human ope a o s by pe o ming au oma ic du ies and le ing
humans handle high-impo ance wo k o es ablish policies and
pe o m o e sigh esponsibili ies. AI suppo s da a analy ical
unc ions while i au oma es adminis a i e wo k alongside se ice
imp o emen p og ams un il human wo ke s main ain hei
posi ions. O ganiza ions equi e ongoing adjus men s along wi h
specialized aining and human-cen e ed e hical s anda ds o
manage AI sys ems wi hin es ablished human-based p inciples.
The model design deli e s ope a ional excellence o public se ices
and p ese es jobs wi hin he bu eauc acy.
Scena io 3: AI Su passes Human Con ol
The mos conce ning si ua ion occu s when au oma ed sys ems
along wi h AI su pass human managemen which igge s an
unp eceden ed le el o au oma ion. The implemen a ion o AI
algo i hms wi hin adminis a ion unc ions leads o mass job losses
along wi h a go e ning sys em con olled by a i icial in elligence.
The powe shi would es ablish AI sys ems wi h con ol ha
exceeds human capabili ies, he eby c ea ing condi ions o
dependency be ween humans and hese AI sys ems. The
dis ibu ion o powe among AI-con olled en i ies while AI makes
decisions and iola es p i acy ele a es e hical isks h oughou
mode n socie y. This scena io c ea es declining human supe ision
and exposes communi ies o elinquish con ol o hei public
adminis a i e ope a ions.
Discussion
These esul s o his esea ch a e discussed wi h ecen ly
esea ches as (C iado & e al, 2024), (Madan & e al, 2023), and
(Kulal & e al, 2024). especially, he esea ches ha o e a de ailed
analysis o he impac ha a i icial in elligence has had on public
adminis a ion. Compa ing and con as ing he indings o he
a ious s udies ha ha e been conduc ed on he subjec o a i icial
in elligence and go e nance is a bene icial way o ha e a be e
unde s anding o he impac ha AI has on go e nance.
When conside ing he exis ing esea ch, i showed ha , (C iado &
e al, 2024) sugges ed a h ee- ie ed amewo k o s udying he
e ec s o AI on public adminis a ion in he yea 2024. This
amewo k would include mac o, meso, and mic o le els. This
p esen esea ch is an ex ension o his app oach by in eg a ing
ac ual da a on AI adop ion ends among U.S. agencies and ac ual
case s udies. As p agma ic implemen a ion p oblems a he agency
le el, e.g., labou eloca ion and issues wi h us , hese indings
sugges , C iado e al. ocused on go e nance-le el measu es.
(Madan & e al, 2023) also in es iga ed AI adop ion ia he p ism
o abso p i e capaci y and public alue ensions. Ou esea ch
alida es hei claim ha go e nance s uc u es and echnology
eadiness shape AI in eg a ion. Bu we also disco e ed ha , gi en
di e ences be ween ede al, s a e, and local go e nmen en i ies
ac oss he Uni ed S a es, he ac ual applica ion o a i icial
in elligence is mo e dispe sed.
Acco ding o he esul s o a ecen s udy on he e ec o a i icial
in elligence on he e ec i eness o public sec o ope a ions by
(Kulal & e al, 2024), AI g ea ly imp o es he ope a ions o
municipal o ganiza ions while ha ing a negligible e ec on
human-cen ic se ices. Ou indings suppo he claim ha
a i icial in elligence-d i en au oma ion inc eases ope a ional
e ec i eness, bu hey also highligh u he ques ions abou how i
a ec s decision-making anspa ency mo e b oadly. Acco ding o
he esul o ou esea ch, ci izens' us in a i icially in elligen -
d i en se ices is lowe han an icipa ed, which con adic s he
esea ch inding o Kulal (Kulal & e al, 2024). I can be seen ha
he e is an u gen need o mo e obus anspa ency and
accoun abili y mechanisms. When compa ed o o he coun ies, he
Uni ed S a es o Ame ica also in es s a subs an ial amoun o
money on he egula ion o a i icial in elligence; i is es ima ed
ha his cos will each up o $9 billion by he yea 2025.
Howe e , ou indings demons a e ha he e is a signi ican
dispa i y be ween eadiness and in es men .
Acco ding o he indings o ou esea ch, ins i u ions in he Uni ed
S a es o Ame ica employ a s a egy ha is no ela ed o any
pa icula app oach o AI go e nance. This is in con as o he
indings o s udies ha ha e been unde aken in Eu ope and India,
which highligh coo dina ed policies o AI go e nance.
Bo h da a go e nance and algo i hmic bias a e conside ed o be
con empo a y p oblems in all he s udies ha ha e been conduc ed
on he subjec .
Copy igh © ISRG Publishe s. All igh s Rese ed.
DOI: 10.5281/zenodo.17283672
152
As pe he esul s o (Madan & e al, 2023) and (C iado & e al,
2024), all ha is needed is esponsibili y and openness. Mo eo e ,
hese pieces o e idence a e complemen a y o one ano he and
unc ion oge he . The e idence ha we p esen , on he o he hand,
p esen s an al e na i e pe spec i e on he deba e: al hough a i icial
in elligence p omo es e iciency, i does no ine i ably inc ease
public us , pa icula ly in ma e s pe aining o he adminis a ion
o law en o cemen and he dis ibu ion o wel a e bene i s. While
he majo i y o he s udy ha was done in he pas concen a ed
p ima ily on he use o a i icial in elligence wi hin go e nmen
s uc u es, he cu en esea ch e eals ha signi ican dis inc ions
a e being made wi h ega d o he deploymen o AI. Despi e he
ac ha law en o cemen agencies a e ha ing ouble dealing wi h
p oblems ela ed o p ejudice, he implemen a ion o a i icial
in elligence in axa ion ope a ions and he igh agains aud has
been success ul o a g ea e deg ee. The p e ious esea ch did no
ully in es iga e his sec o al analysis as i was ca ied ou . Table 5
shows he main di e ences be ween his cu en s udy and ecen ly
esea ches,
Thus, his cu en s udy p esen s a o wa d-looking pe spec i e on
he applica ion o a i icial in elligence in Ame ican public
adminis a ion by combining da a om o he sou ces and
con as ing hem wi h he body o p esen li e a u e. Issues include
employmen loss, lack o openness, and go e nmen al
agmen a ion s ill exis e en i a i icial in elligence has enhanced
pe o mance and decision-making. Fu u e esea ch has o ocus on
de eloping a uni o m go e nance amewo k o a i icial
in elligence i we a e o close he in es men - o- use disc epancy.
Table 5: Compa ison be ween his cu en s udy and he ecen ly esea ches
S udy
Focus A ea
Resul s
Di e ences om Cu en S udy
(C iado & e al,
2024)
AI go e nance
s a egies
F amewo k o AI go e nance a mac o,
meso, and mic o le els
This cu en s udy ex ends o eal-wo ld
agency-le el implemen a ion
(Madan & e al,
2023)
AI and public alue
ensions
AI success depends on abso p i e capaci y
and go e nance amewo ks
The p esen s udy highligh s dispa i ies in
adop ion a di e en go e nmen le els
(Kulal & e al,
2024)
AI in public se ice
e iciency
AI enhances municipal p ocesses bu has
mode a e human-cen ic impac
Ou s udy inds lowe public us han
expec ed
Cu en S udy
AI in U.S. public
adminis a ion
AI imp o es e iciency bu aises
anspa ency and wo k o ce conce ns
The cu en s udy in oduces sec o -speci ic
implemen a ion dispa i ies
AI implemen a ion wi hin go e nmen ins i u ions has simpli ied
adminis a i e p ocesses bu ci izens emain conce ned abou
sys em anspa ency oge he wi h accoun abili y and
disc imina ion isks. AI echnology adop ion ends con inue o
escala e which calls o s eng hened e hical guidelines o manage
hei dis ibu ion in socie y. Public con idence becomes mo e
signi ican because ci izens wan o unde s and be e how AI
sys ems make decisions. The employmen sec o expe iences a
undamen al shi because au oma ion pe o ms ce ain asks ye
de elops esh AI o e sigh and go e nance posi ions.
Public adminis a ion's AI u u e hinges mainly on s a egic
planning oge he wi h egula o y app oaches. Th ee u u e
scena ios a e eme ging om scena io-based analysis ha p esen
AI ei he unde human con ol o alongside humans o domina ing
independen om human in ol emen . The as p og ess o AI
demands egula ly upda ed policies since human o e sigh
ep esen s he bes app oach. Lack o sui able measu es pu s
human con ol a isk which migh c ea e go e nance imbalances
and e hical issues. To gua an ee AI unc ions as a human capabili y
enhance e sus human subs i u e policymake s mus ake
p oac i e s eps while de eloping hei wo k o ce.
Conclusion
Whe e public adminis a ion is conce ned, in eg a ion o a i icial
in elligence p esen s ma elous oppo uni ies and eno mous
challenges. This esea ch s udy es ablishes ha go e nmen
pe o mance in ma e s such as p edic i e modeling, au oma ed
ci izen se ices, and aud de ec ion has been enhanced h ough
echnologies c ea ed h ough a i icial in elligence. Ye he e a e
signi ican obs acles o o e come in he o m o conce ns o e
openness, une en policy en o cemen , and ci izens' lack o us in
a i icial in elligence making decisions. Maybe he mos su p ising
o his s udy is he disc epancy be ween he le el o echnological
eadiness di e en go e nmen agencies ac ually possess and he
inancial in es men in a i icial in elligence. Despi e ede al
agencies ecei ing ample unding o a i icial in elligence, s a e
and local go e nmen s a e ypically no gi en he ools and
expe ise o e ec i ely execu e AI plans. Fu he mo e, e hical
issues su ounding algo i hmic bias and da a p i acy unde sco e
he necessi y o c a ing obus laws o go e n he up ake o
a i icial in elligence. Subsequen esea ch needs o cen e mainly
on he de elopmen o s anda dized models o go e nance and
longi udinal s udies quan i ying he e ec o a i icial in elligence
egula ion in he eal wo ld. A i icial in elligence is on a as
ack. The policymake s should ensu e ha a i icial in elligence
is an asse o imp o ed public adminis a ion and no a cause o
inequali y and mis us ; hus, hey should c ea e maximum p io i y
owa ds an app oach weighing he ad an ages o AI agains i s
na u al isks.
Recommenda ions
The ollowing ecommenda ions se e as ecommenda ions o
handle he challenges oge he wi h he isks ha AI echnology
p esen s o public adminis a ion:
Es ablish clea policies oge he wi h e hical guidelines
which main ain AI as a ool helping people no eplacing
hei asks.
Copy igh © ISRG Publishe s. All igh s Rese ed.
DOI: 10.5281/zenodo.17283672
153
Public sec o o ganiza ions need o design aining
schemes ha de elop wo k o ce capabili ies o wo king
wi h AI-based sys ems.
Go e nmen s should implemen p ocedu es o
de eloping AI which main ains bo h anspa ency and
accoun abili y o a oiding bias and ensu ing ai
go e nance p ac ices.
The coexis ence be ween humans and machines should
be sus ained h ough policies which guide AI au oma ion
p ac ices while keeping human supe ision ac i e.
AI Accoun abili y Sys ems need de elopmen o
acking AI sys ems ha e ain om c ucial choices
when pe sonnel should be in ol ed.
Disclosu e S a emen
E hical app o al and consen o pa icipa e: Conduc ed in
acco dance wi h es ablished p ocedu es.
A ailabili y o da a and ma e ials: Collec ed in a p ope and
sys ema ic manne .
Au ho con ibu ion: The au ho was solely esponsible o all
s ages o he esea ch and manusc ip w i ing.
Con lic o in e es : The au ho has no con lic s o in e es o
disclose.
Funding: No unding o inancial suppo was ecei ed om any
o ganiza ion.
Acknowledgmen s: The esea che ex ends hea el hanks o all
iends and colleagues who con ibu ed o he success o his wo k,
and since e app ecia ion o he dea s uden s who pa icipa ed in
he s udy.
Open Access
This a icle is licensed unde a C ea i e Commons A ibu ion 4.0
In e na ional License, which pe mi s use, sha ing, adap a ion,
dis ibu ion and ep oduc ion in any medium o o ma , as long as
you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce,
p o ide a link o he C ea i e Commons licence, and indica e i
changes we e made. The images o o he hi d pa y ma e ial in his
a icle a e included in he a icle's C ea i e Commons licence,
unless indica ed o he wise in a c edi line o he ma e ial. I
ma e ial is no included in he a icle's C ea i e Commons licence
and you in ended use is no pe mi ed by s a u o y egula ion o
exceeds he pe mi ed use, you will need o ob ain pe mission
di ec ly om he copy igh holde . To iew a copy o his license,
isi h ps://c ea i ecommons.o g/licenses/by-nc/4.0/
Re e ences
1. Adewusi, & e al. (2024). A i icial in elligence in
cybe secu i y: P o ec ing na ional in as uc u e: A USA.
Wo ld Jou nal o Ad anced Resea ch and Re iews,
2263-2275.
2. Agba, & e al. (2023). A i icial in elligence and public
managemen and go e nance in de eloped and
de eloping ma ke economies. Jou nal o Public
Adminis a ion, Policy and Go e nance Resea ch, 1-14.
3. Babšek, & e al. (2025). A i icial In elligence Adop ion
in Public Adminis a ion: An O e iew o Top-Ci ed
A icles and P ac ical Applica ions. AI, 44.
4. Bullock, & e al. (2020). A i icial in elligence,
bu eauc a ic o m, and disc e ion in public se ice.
In o ma ion Poli y, 491-506.
5. Caiza, & e al. (2024). Na iga ing Go e nmen al
Choices: A Comp ehensi e Re iew o A i icial
In elligence’s Impac on Decision-Making. MDPI, 64.
6. Cen e s o Medica e & Medicaid Se ices. (2019).
F aud P e en ion and AI-Based Decision Sys ems.
Go e nmen Repo .
7. Chen, & e al. (2024). The adop ion and implemen a ion
o a i icial in elligence cha bo s in public o ganiza ions:
E idence om US s a e go e nmen s. The Ame ican
Re iew o Public Adminis a ion, 255-270.
8. C iado, & e al. (2024). A i icial in elligence and public
adminis a ion: Unde s anding ac o s, go e nance, and
policy om mic o, meso, and mac o pe spec i es. Public
Policy and Adminis a ion, 09520767241272921.
9. Fosso Wamba, & e al. (2024). Cha GPT and gene a i e
a i icial in elligence: an explo a o y s udy o key
bene i s and challenges in ope a ions and supply chain
managemen . In e na ional Jou nal o P oduc ion
Resea ch, 5676-5696.
10. He mann, & e al. (2024). Deploying a i icial
in elligence in se ices o AID ulne able consume s.
Jou nal o he Academy o Ma ke ing Science, 1431-
1451.
11. Ins i u e, A. N. (2019). Au oma ed Decision Sys ems in
US Go e nmen . Resea ch Repo .
12. Kulal, & e al. (2024). Enhancing public se ice deli e y
e iciency: Explo ing he impac o AI. Jou nal o Open
Inno a ion: Technology, Ma ke , and Complexi y,
100329.
13. Madan, & e al. (2023). AI adop ion and di usion in
public adminis a ion: A sys ema ic li e a u e e iew and
u u e esea ch agenda. Go e nmen In o ma ion
Qua e ly, 101774.
14. Mohammed, & e al. (2024). The ole o a i icial
in elligence and in ech in p omo ing eco- iendly
in es men s and non-g eenwashing p ac ices in he US
ma ke . Jou nal o En i onmen al Managemen , 120977.
15. Nade , & e al. (2024). Public unde s anding o a i icial
in elligence h ough en e ainmen media. AI & socie y,
713-726.
16. Na ional Academy o Public Adminis a ion. (2019). AI
in Go e nmen : Policies and Implemen a ion S a egies.
NAPA Repo .
17. Schi , & e al. (2022). Assessing public alue ailu e in
go e nmen adop ion o a i icial in elligence. Public
Adminis a ion, 653-673.
18. Schwei ze , B. (2024). A i icial in elligence (AI) e hics
in accoun ing. Jou nal o Accoun ing, E hics & Public
Policy, 67-67.
19. Uzun, & e al. (2022). Big Ques ions o AI in Public
Adminis a ion and Policy. Siyasal: Jou nal o Poli ical
Sciences, 423-442.
20. Wang, & e al. (2024). Balancing inno a ion and
Regula ion in he age o geneRa i e a i icial
in elligence. Jou nal o In o ma ion Policy, 14.
