scieee Science in your language
[en] (orig)

Zero trust in cloud infrastructure: Implementing secure CI/CD Pipelines

Author: Kadulla, Sumanth
Publisher: Zenodo
DOI: 10.5281/zenodo.17292231
Source: https://zenodo.org/records/17292231/files/WJARR-2025-1662.pdf
 Co esponding au ho : Suman h Kadulla
Copy igh © 2025 Au ho (s) e ain he copy igh o his a icle. This a icle is published unde he e ms o he C ea i e Commons A ibu ion License 4.0.
Ze o us in cloud in as uc u e: Implemen ing secu e CI/CD Pipelines
Suman h Kadulla *
Wes e n Illinois Uni e si y, USA.
Wo ld Jou nal o Ad anced Resea ch and Re iews, 2025, 26(02), 450-457
Publica ion his o y: Recei ed on 27 Ma ch 2025; e ised on 03 May 2025; accep ed on 05 May 2025
A icle DOI: h ps://doi.o g/10.30574/wja .2025.26.2.1662
Abs ac
Ze o T us a chi ec u e ep esen s a undamen al shi in secu ing cloud in as uc u e, pa icula ly wi hin CI/CD
pipelines whe e adi ional pe ime e -based secu i y app oaches inc easingly ail agains sophis ica ed h ea s. This
echnical a icle explo es how implemen ing Ze o T us p inciples—"ne e us , always e i y"—c ea es obus
p o ec ion h oughou he so wa e deli e y li ecycle. The implemen a ion spans ac oss mul iple dimensions: secu ing
mode n CI/CD ools including Gi Hub Ac ions, Azu e De Ops, and Gi Lab; es ablishing comp ehensi e iden i y and
access managemen wi h jus -in- ime p i ileged access; le e aging Powe Shell o secu i y au oma ion; and ensu ing
obus con aine secu i y ac oss Docke and Kube ne es en i onmen s. Each dimension con ibu es o a de ense-in-
dep h s a egy ha add esses he unique challenges o cloud-na i e en i onmen s. The a icle demons a es how
explici e i ica ion o all access eques s, ega dless o o igin, combined wi h ine-g ained pe missions, con inuous
moni o ing, and au oma ed compliance alida ion c ea es signi ican ly enhanced secu i y pos u es. Fo o ganiza ions
unde going digi al ans o ma ion wi h au oma ed so wa e deli e y pipelines, adop ing hese Ze o T us
me hodologies ensu es applica ion in eg i y h oughou he de elopmen li ecycle while main aining he agili y bene i s
ha make cloud en i onmen s aluable in he i s place.
Keywo ds: Ze o T us A chi ec u e; CI/CD Secu i y; Cloud In as uc u e; Con aine O ches a ion; Iden i y
Managemen
1. In oduc ion
In oday's apidly e ol ing cloud landscape, secu i y can no longe be an a e hough . The adi ional pe ime e -based
secu i y model has p o en inadequa e agains sophis ica ed cybe h ea s ha can comp omise sys ems om bo h
ex e nal and in e nal ec o s. A comp ehensi e su ey o secu i y p o essionals e ealed ha 83% ha e expe ienced
secu i y inciden s ela ed o hei cloud se ices, wi h 41% epo ing ha hese inciden s speci ically a ge ed hei
CI/CD pipelines o de elopmen in as uc u e [1]. Ze o T us a chi ec u e has eme ged as a compelling al e na i e,
ope a ing on he p inciple o "ne e us , always e i y." This app oach is pa icula ly c ucial in cloud in as uc u e,
whe e he dynamic na u e o esou ces and dis ibu ed en i onmen s c ea es complex secu i y challenges.
The su ey also ound ha 72% o o ganiza ions a e now p io i izing AI-based secu i y solu ions o add ess he
inc easing complexi y o h ea s in cloud en i onmen s, wi h 68% speci ically implemen ing hese echnologies o
p o ec hei De Ops wo k lows. Mos conce ning, 57% o esponden s indica ed ha adi ional secu i y con ols we e
ine ec i e agains sophis ica ed a acks a ge ing hei de elopmen pipelines.
Recen analysis o ac ual da a b eaches e ealed ha o ganiza ions implemen ing Ze o T us p inciples educed he
a e age cos o a da a b each by 2.2 million USD compa ed o hose elying on adi ional secu i y models [2]. Examining
553 o ganiza ions ac oss mul iple coun ies and indus ies demons a ed ha he a e age cos o a da a b each in cloud
en i onmen s eached 4.99 million USD in 2024 o o ganiza ions wi hou comp ehensi e Ze o T us con ols,
Wo ld Jou nal o Ad anced Resea ch and Re iews, 2025, 26(02), 450-457
451
ep esen ing a 17% inc ease om he p e ious yea [2]. Fu he mo e, o ganiza ions wi h ma u e cloud secu i y
p ac ices expe ienced 52 ewe days o ope a ional dis up ion ollowing a b each compa ed o hose wi h less
de eloped secu i y amewo ks.
This a icle explo es he implemen a ion o Ze o T us p inciples wi hin cloud-based CI/CD pipelines. As o ganiza ions
accele a e hei digi al ans o ma ion ini ia i es, he au oma ion o so wa e deli e y h ough CI/CD pipelines has
become s anda d p ac ice, wi h indus y analysis showing ha 78% o en e p ise o ganiza ions ha e adop ed some
o m o De Ops p ac ices inco po a ing CI/CD me hodologies. Howe e , hese pipelines ha e become a ac i e a ge s
o a acke s seeking o injec malicious code o exploi ulne abili ies in he deploymen p ocess.
The secu i y landscape has become pa icula ly conce ning as 63% o o ganiza ions epo expe iencing dis up ions o
c i ical in as uc u e ollowing secu i y inciden s, wi h an a e age eco e y ime o 108 days o he mos se e e
b eaches [2]. By adop ing Ze o T us me hodologies, o ganiza ions can es ablish obus secu i y con ols ha e i y
e e y access eques ega dless o o igin, ensu ing he in eg i y o applica ions h oughou he de elopmen li ecycle.
This app oach becomes essen ial as 74% o su eyed secu i y p o essionals belie e ha AI-powe ed a acks agains
cloud in as uc u e and CI/CD pipelines will inc ease signi ican ly in complexi y and equency o e he nex wo yea s
[1].
Figu e 1 Cloud Secu i y Inciden s [1, 2]
2. Unde s anding Ze o T us A chi ec u e in Cloud En i onmen s
2.1. Co e P inciples o Ze o T us
Ze o T us a chi ec u e undamen ally ans o ms secu i y app oaches in cloud en i onmen s by elimina ing implici
us and con inuously alida ing e e y digi al in e ac ion. Acco ding o NIST's Ze o T us A chi ec u e amewo k,
o ganiza ions ha ully implemen he esou ce-cen ic e i ica ion model expe ienced 43% ewe success ul b each
a emp s in he p e ious iscal yea [3]. This amewo k iden i ies h ee co e p inciples ha d i e meaning ul secu i y
ans o ma ion.
Ve i y explici ly demands con inuous au hen ica ion and au ho iza ion based on comp ehensi e da a poin s. A s udy o
1,385 cloud secu i y implemen a ions ound ha o ganiza ions employing con inuous e i ica ion de ec ed
unau ho ized access a emp s 5.2 imes as e han hose using pe iodic au hen ica ion. The NIST app oach speci ically
ad oca es o secu i y pos u e e i ica ion o all se en en e p ise esou ces: iden i y, de ice, ne wo k/en i onmen ,
applica ion wo kload, da a, isibili y/analy ics, and go e nance [3]. Analysis shows o ganiza ions implemen ing
e i ica ion ac oss all se en domains educed hei ulne abili y o c eden ial-based a acks by 76%.
Wo ld Jou nal o Ad anced Resea ch and Re iews, 2025, 26(02), 450-457
452
Leas p i ileged access emains cen al o e ec i e Ze o T us implemen a ion. Acco ding o NIST guidance, ma u e
o ganiza ions educed excessi e p i ileges by implemen ing dynamic access con ols ha limi no only who can access
esou ces bu also wha ac ions can be pe o med, om whe e, and o how long [3]. This app oach esul ed in 67%
ewe p i ilege escala ion inciden s compa ed o s a ic models.
Assume b each shi s ocus om pe ime e de ense o in e nal moni o ing. NIST guidance indica es ha o ganiza ions
ea ing all ne wo k a ic as po en ially malicious de ec ed ac ual b eaches 57% as e han p e en ion- ocused
app oaches [3]. By limi ing implici us zones and implemen ing comp ehensi e session-based au hen ica ion,
o ganiza ions educed la e al mo emen in 81% o simula ed a acks.
2.2. Challenges in T adi ional CI/CD Secu i y Models
T adi ional CI/CD secu i y models ace subs an ial limi a ions in mode n cloud en i onmen s. Recen esea ch
examining 375 de elopmen pipelines iden i ied c i ical secu i y gaps ha Ze o T us p inciples di ec ly add ess [4].
T adi ional app oaches ely hea ily on pe ime e p o ec ion ha ails o coun e sophis ica ed h ea s.
Inside h ea s om p i ileged use s ep esen a signi ican challenge, wi h esea ch indica ing 39% o cloud secu i y
inciden s in ol ed c eden ial misuse by au ho ized pe sonnel. In es iga ion o comp omised pipelines e ealed 83%
in ol ed c eden ials wi h unnecessa y ele a ed pe missions [4].
Supply chain secu i y emains pa icula ly p oblema ic, wi h 52% o o ganiza ions epo ing di icul y main aining
isibili y ac oss hei de elopmen dependencies [4]. The a e age en e p ise pipeline inco po a es 127 ex e nal
componen s, wi h comp ehensi e secu i y alida ion pe o med on only 36% o hese dependencies.
2.3. Bene i s o Ze o T us in CI/CD Pipelines
Implemen ing Ze o T us p inciples deli e s measu able secu i y imp o emen s. O ganiza ions adop ing
comp ehensi e e i ica ion models o hei de elopmen wo k lows epo 62% educ ion in success ul a acks
a ge ing deli e y in as uc u e [4].
Con inuous e i ica ion subs an ially imp o es secu i y pos u e, wi h esea ch showing 79% educ ion in unau ho ized
code inse ions. Indus y analysis indica es ha e i ica ion ga es h oughou he pipeline can p e en 91% o po en ial
comp omise a emp s while adding minimal ope a ional o e head [4].
Ze o T us implemen a ion imp o es egula o y compliance ou comes by 51%, wi h o ganiza ions epo ing
signi ican ly educed indings du ing o mal assessmen s [4]. The amewo k's emphasis on con inuous moni o ing and
explici e i ica ion aligns wi h majo compliance equi emen s, educing duplica i e secu i y con ols while
s eng hening o e all pos u e.
Table 1 Measu able bene i s o implemen ing Ze o T us a chi ec u e [3, 4]
Bene i
Imp o emen Pe cen age
B each Cos Reduc ion
$2.2M USD
B each De ec ion Speed
57% as e
La e al Mo emen P e en ion
81% educ ion
C eden ial A ack Vulne abili y
76% educ ion
P i ilege Escala ion Inciden s
67% educ ion
Success ul B each A emp s
43% educ ion
3. Implemen ing Secu e CI/CD Pipelines wi h Mode n Tools
3.1. Gi Hub Ac ions Secu i y Bes P ac ices
Gi Hub Ac ions has eme ged as a leading pla o m o CI/CD au oma ion, bu i s powe ul capabili ies necessi a e obus
secu i y measu es. Acco ding o ecen indus y esea ch, 67% o o ganiza ions using au oma ed pipelines expe ienced
Wo ld Jou nal o Ad anced Resea ch and Re iews, 2025, 26(02), 450-457
453
a leas one secu i y inciden ela ed o miscon igu a ions [5]. Comp ehensi e secu i y p ac ices a e essen ial o
main aining pipeline in eg i y.
Reposi o y and o ganiza ion-le el pe missions o m he ounda ion o Gi Hub Ac ions secu i y. Resea ch indica es ha
o ganiza ions implemen ing ine-g ained pe missions expe ienced 61% ewe unau ho ized access inciden s. By
es ic ing wo k low execu ion igh s o speci ic eams, secu i y eams educed he po en ial a ack su ace signi ican ly.
A no able inding shows ha 83% o secu i y inciden s o igina ed om excessi e pe missions, unde sco ing he
impo ance o p ope access con ols.
Secu ing wo k low iles h ough sys ema ic YAML alida ion subs an ially educes miscon igu a ions. Analysis e eals
ha 45% o o ganiza ions ha implemen ed au oma ed secu i y scanning disco e ed c i ical ulne abili ies ha would
ha e o he wise eached p oduc ion [5]. O ganiza ions implemen ing code scanning as pa o hei pull eques p ocess
p e en ed 85% o high- isk ulne abili ies om en e ing he codebase.
Sec e managemen equi es pa icula a en ion, wi h esea ch indica ing ha exposed sec e s con ibu ed o 38% o
success ul a acks. O ganiza ions le e aging dedica ed sec e s managemen educed c eden ial exposu e by 72%. The
indings sugges ha 91% o o ganiza ions a e s ill s o ing sec e s as plain ex in some pa s o hei sys ems, c ea ing
subs an ial secu i y isks [5].
B anch p o ec ion and equi ed app o als ep esen c i ical secu i y con ols. Resea ch demons a es ha eposi o ies
wi h p o ec ion ules expe ience 76% ewe unau ho ized modi ica ions. O ganiza ions implemen ing equi ed e iews
o p oduc ion wo k lows educed malicious code inse ions by 89%. The da a shows ha p ope e iew p ocesses
de ec ed 79% o po en ial secu i y issues be o e hey en e ed p oduc ion code.
3.2. Azu e De Ops Pipeline Secu i y
Resea ch ac oss en e p ise De Ops deploymen s e ealed ha o ganiza ions implemen ing a s uc u ed ma u i y
model o secu i y con ols expe ienced 64% ewe success ul a acks [6]. Acco ding o he De SecOps Ma u i y Model
analysis, only 34% o o ganiza ions ha e eached le el 3 o highe ma u i y in hei pipeline secu i y implemen a ion.
Se ice connec ions wi h managed iden i ies deli e subs an ial secu i y bene i s. Analysis indica es ha o ganiza ions
implemen ing au oma ed c eden ial managemen educed secu i y inciden s by 71%. This app oach minimized he
human e o ac o ha con ibu ed o 53% o c eden ial exposu es [6].
Pipeline en i onmen s wi h app o al ga es c ea e e ec i e secu i y bounda ies. O ganiza ions implemen ing
seg ega ed en i onmen s wi h explici app o als educed unau ho ized deploymen s by 82%. The ma u i y model da a
shows ha 57% o o ganiza ions s ill s uggle wi h p ope en i onmen isola ion, c ea ing subs an ial la e al mo emen
oppo uni ies o a acke s [6].
3.3. Gi Lab CI/CD Secu i y Con ols
Analysis o en e p ise CI/CD implemen a ions e ealed ha o ganiza ions ollowing a s uc u ed ma u i y model
expe ienced 59% ewe success ul a acks [6]. The esea ch indica es ha only 28% o o ganiza ions ha e achie ed
le el 4 ma u i y in unne isola ion and execu ion con ol.
Secu e a iables and en i onmen sepa a ion c ea e essen ial secu i y bounda ies. Acco ding o he ma u i y model
assessmen , o ganiza ions implemen ing dis inc en i onmen con igu a ions educed c eden ial exposu e by 77%. The
s udy e ealed ha 61% o o ganiza ions ail o implemen p ope c eden ial managemen , wi h a iable scope e o s
con ibu ing o 43% o secu i y inciden s [6].
Wo ld Jou nal o Ad anced Resea ch and Re iews, 2025, 26(02), 450-457
454
Figu e 2 E ec i eness o secu i y con ols ac oss di e en CI/CD pla o ms [5, 6]
4. Iden i y and Access Managemen in Cloud CI/CD
4.1. IAM Roles and Jus -in-Time Access
Iden i y and Access Managemen se es as he co ne s one o Ze o T us implemen a ion in cloud CI/CD en i onmen s.
Resea ch e eals ha 99% o cloud iden i ies use less han 5% o he pe missions g an ed o hem, c ea ing subs an ial
unnecessa y isk [7]. This excessi e pe mission gap ep esen s a c i ical ulne abili y in mos CI/CD pipelines.
Fine-g ained ole-based access con ol d ama ically educes he a ack su ace. The S a e o Cloud Pe missions epo
ound ha 90% o o ganiza ions ha e iden i ies wi h high- isk pe missions hey' e ne e used, ye hese pe missions
emain ac i e [7]. When o ganiza ions implemen leas -p i ilege p inciples h ough g anula RBAC, hey see
measu able secu i y imp o emen s. Analysis indica es ha he a e age iden i y main ains access o 10-20 imes mo e
esou ces han ac ually needed o ope a ional unc ionali y.
Tempo a y c eden ial issuance h ough sho -li ed okens subs an ially enhances secu i y pos u e. Resea ch indica es
ha iden i ies wi h access o sensi i e da a expe ience 18 imes mo e pe mission misuse han hose wi h s anda d
access le els [7]. Implemen a ions using okens wi h maximum li espans o 4-8 hou s p e en mos la e al mo emen
a acks ha succeed agains sys ems wi h pe sis en c eden ials.
Jus -in- ime p i ileged access managemen ans o ms adi ional pe mission models by p o iding ele a ed access only
when needed. Da a shows ha 62% o o ganiza ions ail o implemen a o mal li ecycle o cloud pe missions, esul ing
in "pe mission sp awl" ha g ows by app oxima ely 25% each qua e [7]. By implemen ing JIT access wi h au oma ed
app o al wo k lows, o ganiza ions educe he p i ileged access window while main aining ope a ional e ec i eness.
4.2. Powe Shell Au oma ion o Secu e Access Con ol
Powe Shell au oma ion p o ides essen ial capabili ies o implemen ing and main aining obus secu i y con ols.
Resea ch indica es ha au oma ion educes secu i y con igu a ion e o s by app oxima ely 70% while educing
deploymen imes by up o 90% [8]. Au oma ion becomes inc easingly c i ical as en i onmen s g ow in complexi y.
Sec e and c eden ial o a ion au oma ion signi ican ly educes bo h secu i y isks and ope a ional o e head. Analysis
shows ha implemen ing au oma ed o a ion educes a e age c eden ial li espans by 75% while dec easing
adminis a i e e o signi ican ly [8]. The in eg a ion o au oma ed secu i y checks in o he CI/CD pipeline ensu es
consis en secu i y p ac ices ac oss all deploymen s.

Wo ld Jou nal o Ad anced Resea ch and Re iews, 2025, 26(02), 450-457
455
Cus om secu i y modules aligned o o ganiza ional policies ensu e consis en implemen a ion ac oss di e se
en i onmen s. Da a shows ha s anda dized secu i y au oma ion enables in as uc u e eams o spend 60% less ime
on secu i y- ela ed asks while imp o ing compliance me ics [8]. By embedding secu i y in o au oma ion p ocesses,
o ganiza ions es ablish con inuous secu i y alida ion h oughou he deploymen li ecycle.
4.3. C eden ial Managemen and Ro a ion
Robus c eden ial managemen o ms a c i ical ounda ion o Ze o T us implemen a ion. Resea ch indica es ha 83%
o o ganiza ions lack p ope acking o which iden i ies ha e access o sensi i e esou ces [7]. P ope c eden ial
go e nance becomes especially c i ical conside ing ha sensi i e da a access ca ies subs an ially highe isk o misuse.
Au oma ed c eden ial o a ion schedules d as ically educe secu i y isks associa ed wi h s a ic c eden ials. Analysis
shows ha o ganiza ions implemen ing egula o a ion educe he po en ial impac window o comp omised
c eden ials by o e 70% [8]. The implemen a ion o cen alized managemen wi h au oma ed o a ion p e en s many
common c eden ial-based a ack ec o s while imp o ing ope a ional eliabili y.
5. Con aine iza ion and O ches a ion Secu i y
5.1. Docke Con aine Secu i y
Con aine s ha e ans o med applica ion deploymen , bu in oduce unique secu i y challenges equi ing sys ema ic
app oaches. Indus y esea ch indica es a signi ican majo i y o o ganiza ions now use con aine s in p oduc ion, wi h
o e hal expe iencing secu i y inciden s in he pas yea [9]. As con aine adop ion accele a es, comp ehensi e secu i y
p ac ices become essen ial.
Minimal base images and mul i-s age builds p o ide ounda ional secu i y bene i s. The con aine secu i y guide
emphasizes ha con aine images should ollow he p inciple o leas unc ionali y, including only wha 's necessa y o
un he applica ion [9]. By sepa a ing build en i onmen s om un ime en i onmen s, o ganiza ions signi ican ly
educe a ack su aces while imp o ing ope a ional e iciency.
Regula ulne abili y scanning h oughou he con aine li ecycle o e s essen ial isibili y in o secu i y isks. The
con aine secu i y amewo k ecommends implemen ing scanning a mul iple phases: du ing de elopmen , be o e
pushing o egis ies, and con inuously in un ime en i onmen s [9]. This mul i-laye ed app oach enables ea ly
de ec ion o ulne abili ies ac oss he en i e so wa e supply chain.
Image signing es ablishes c yp og aphic e i ica ion o con aine au hen ici y. The con aine secu i y bes p ac ices
highligh ha image signing c ea es a chain o us om de elopmen h ough p oduc ion deploymen [9]. This
alida ion p e en s unau ho ized modi ica ions ha could in oduce malicious code o backdoo s in o he deploymen
pipeline.
5.2. Kube ne es Secu i y Pos u e
Kube ne es secu i y equi es comp ehensi e de ensi e measu es ac oss mul iple laye s. Secu i y esea che s no e ha
Kube ne es p esen s signi ican challenges due o i s complex a chi ec u e and ex ensi e a ack su ace [10]. As
adop ion g ows, o ganiza ions mus implemen sys ema ic p o ec ion.
Pod secu i y policies c ea e c i ical secu i y bounda ies wi hin clus e s. The Kube ne es secu i y model emphasizes
es ic ing p i ileges, p e en ing sensi i e moun poin s, and en o cing ead-only oo ilesys ems [10]. These
p o ec ions signi ican ly educe he impac o con aine comp omise while main aining ope a ional unc ionali y.
Role-based access con ol implemen a ion p e en s unau ho ized adminis a i e ac ions. Secu i y bes p ac ices s ess
ha RBAC should ollow leas -p i ilege p inciples, wi h explici pe mission g an s a he han b oad access [10]. The
p inciple o leas p i ilege applies a e e y le el: use s, applica ions, and in as uc u e componen s.
5.3. AKS and OpenShi Secu i y Fea u es
Managed Kube ne es pla o ms p o ide enhanced secu i y capabili ies add essing common o ches a ion challenges.
Acco ding o secu i y esea ch, hese pla o ms in eg a e nume ous p o ec ions ha would o he wise equi e
signi ican con igu a ion e o [10].
Wo ld Jou nal o Ad anced Resea ch and Re iews, 2025, 26(02), 450-457
456
In eg a ion wi h cloud p o ide iden i y se ices es ablishes consis en au hen ica ion ac oss en i onmen s. Secu i y
amewo ks emphasize he impo ance o cen alized iden i y managemen wi h ede a ion capabili ies [10]. This
app oach elimina es he need o se ice accoun p oli e a ion while enabling ine-g ained access con ol.
5.4. Compliance and Audi in Con aine O ches a ion
Main aining compliance in con aine ized en i onmen s equi es sys ema ic app oaches add essing scale and
complexi y. The con aine secu i y amewo k emphasizes ha au oma ed compliance alida ion should be in eg a ed
h oughou he de elopmen and deploymen li ecycle [9].
Con inuous compliance alida ion ensu es consis en policy en o cemen . Secu i y bes p ac ices ecommend
embedding policy checks di ec ly in o CI/CD pipelines o p e en non-complian wo kloads om eaching p oduc ion
[9]. This shi -le app oach iden i ies issues ea lie when emedia ion cos s a e signi ican ly lowe .
Table 2 Laye ed secu i y app oaches o con aine en i onmen s [9, 10]
Secu i y Implemen a ion
Secu i y Laye
Minimal Base Images
Con aine Build
Mul i-s age Builds
Con aine Build
Vulne abili y Scanning
De elopmen , Regis y, Run ime
Image Signing
Supply Chain
Pod Secu i y Policies
Clus e Con igu a ion
RBAC Implemen a ion
Access Con ol
Managed Kube ne es In eg a ion
Pla o m
Compliance Au oma ion
De SecOps In eg a ion
6. Conclusion
Ze o T us a chi ec u e has eme ged as an essen ial secu i y pa adigm o p o ec ing mode n cloud in as uc u e,
pa icula ly wi hin CI/CD pipelines ha ha e become bo h c i ical asse s and a ac i e a ack a ge s. The
comp ehensi e implemen a ion o "ne e us , always e i y" p inciples undamen ally ans o ms secu i y ac oss
mul iple dimensions o he deploymen pipeline. By equi ing con inuous e i ica ion o e e y digi al in e ac ion
ega dless o o igin, o ganiza ions es ablish obus p o ec ions ha adi ional pe ime e -based app oaches canno
achie e. The laye ed implemen a ion begins wi h secu ing CI/CD ools like Gi Hub Ac ions, Azu e De Ops, and Gi Lab
h ough ine-g ained pe missions, wo k low alida ion, and sec e managemen . I ex ends h ough iden i y and access
managemen wi h jus -in- ime p i ileges and sho -li ed c eden ials, signi ican ly educing he a ack su ace
h oughou he in as uc u e. Powe Shell au oma ion c ea es consis ency while educing human e o in secu i y
con igu a ions. Con aine secu i y om Docke h ough Kube ne es es ablishes secu e ounda ions wi h minimal base
images, ulne abili y scanning, and c yp og aphic e i ica ion. The secu i y bounda ies ex end h ough pod policies,
ole-based con ols, and managed Kube ne es secu i y ea u es. Au oma ed compliance alida ion h oughou he
pipeline ensu es consis en en o cemen o secu i y s anda ds. Toge he , hese laye s c ea e de ense-in-dep h ha
main ains he in eg i y o applica ions h oughou hei li ecycle. As cloud en i onmen s and h ea s con inue e ol ing,
his Ze o T us app oach becomes inc easingly i al o o ganiza ions seeking o p o ec hei digi al asse s while
p ese ing he agili y bene i s o cloud au oma ion.
Re e ences
[1] Cloud Secu i y Alliance, "The S a e o AI and Secu i y Su ey Repo ," 2024. [Online]. A ailable:
h ps://cloudsecu i yalliance.o g/a i ac s/ he-s a e-o -ai-and-secu i y-su ey- epo
[2] IBM, "IBM Repo : Escala ing Da a B each Dis up ion Pushes Cos s o New Highs," 2024. [Online]. A ailable:
h ps://news oom.ibm.com/2024-07-30-ibm- epo -escala ing-da a-b each-dis up ion-pushes-cos s- o-new-
highs
Wo ld Jou nal o Ad anced Resea ch and Re iews, 2025, 26(02), 450-457
457
[3] Lau en Koppelman, "Wha is he NIST Ze o T us A chi ec u e?," Nex DLP, 2024. [Online]. A ailable:
h ps://www.nex dlp.com/ esou ces/blog/nis -ze o- us -a chi ec u e
[4] SEATTLE, "La es De SecOps Guidance om Cloud Secu i y Alliance and SAFECode Emphasizes Value o
Collabo a ion, In eg a ion in De SecOps Landscape," CSA, 2024. [Online]. A ailable:
h ps://cloudsecu i yalliance.o g/p ess- eleases/2024/02/21/la es -de secops-guidance- om-cloud-
secu i y-alliance-and-sa ecode-emphasizes- alue-o -collabo a ion-in eg a ion-in-de secops-landscape
[5] Resou ces, "E ol ing Gi Hub Ad anced Secu i y: G ea e lexibili y, easie o access," 2025. [Online]. A ailable:
h ps:// esou ces.gi hub.com/e ol ing-gi hub-ad anced-secu i y/
[6] Chai ali Dho e and Abby Taylo , "Building a De SecOps Ma u i y Model: A Roadmap o En e p ises," Qen elli.
[Online]. A ailable: h ps://qen elli.com/ hough -leade ship/insigh s/building-a-de secops-ma u i y-model-a-
oadmap- o -en e p ises
[7] Alex Simons, "2023 S a e o Cloud Pe missions Risks epo now published," Tech Communi y, 2023. [Online].
A ailable: h ps:// echcommuni y.mic oso .com/blog/mic oso -en a-blog/2023-s a e-o -cloud-pe missions-
isks- epo -now-published/1061397
[8] DuploCloud, "Building a Secu e Cloud In as uc u e wi h De Ops," 2025. [Online]. A ailable:
h ps://duplocloud.com/blog/secu e-cloud-in as uc u e-wi h-de ops/
[9] Tige a, "Con aine Secu i y: 7 Key Componen s and 8 C i ical Bes P ac ices." [Online]. A ailable:
h ps://www. ige a.io/lea n/guides/con aine -secu i y-bes -p ac ices/
[10] Rags S ini as, e al., "Kube ne es Secu i y: The S a e o he Union - a Vi ual Panel," In oQ, 2020. [Online].
A ailable: h ps://www.in oq.com/a icles/Kube ne es-secu i y/