Ayinaddis, Samuel Godadaw
A icle
A i icial in elligence adop ion dynamics and knowledge in
SMEs and la ge i ms: A sys ema ic e iew and bibliome ic
analysis
Jou nal o Inno a ion & Knowledge (JIK)
P o ided in Coope a ion wi h:
Else ie
Sugges ed Ci a ion: Ayinaddis, Samuel Godadaw (2025) : A i icial in elligence adop ion dynamics
and knowledge in SMEs and la ge i ms: A sys ema ic e iew and bibliome ic analysis, Jou nal o
Inno a ion & Knowledge (JIK), ISSN 2444-569X, Else ie , Ams e dam, Vol. 10, Iss. 3, pp. 1-15,
h ps://doi.o g/10.1016/j.jik.2025.100682
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/327583
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/
A i icial in elligence adop ion dynamics and knowledge in SMEs and la ge
i ms: A sys ema ic e iew and bibliome ic analysis
Samuel Godadaw Ayinaddis
*
Depa men o Economics and Managemen , Uni e si y o Pisa, I aly
ARTICLE INFO
JEL Classi ica ion:
L22
L25
L26
O30-O33
M15
Keywo ds:
Small and medium en e p ise (SMEs)
A i icial in elligence (AI)
La ge i ms
AI adop ion
ABSTRACT
A i icial in elligence (AI) has quickly eme ged as a op echnological p io i y o companies in a ious sec o s,
adically al e ing business ope a ions. Howe e , he exis ing li e a u e e eals a agmen ed and inconsis en
unde s anding o AI adop ion dynamics be ween small and medium en e p ises (SMEs) and la ge , well-
es ablished i ms. This dicho omy o he exis ing esea ch aises impo an ques ions abou whe he he AI
ools and applica ion modali ies used by hese companies a e inhe en ly simila o i signi ican di e ences exis
in hei implemen a ion and ou comes due o a ying o ganiza ional sizes. This s udy e alua es whe he small
and la ge i ms’ e o s owa d implemen ing AI di e signi ican ly using bibliome ic analysis and a sys ema ic
li e a u e e iew om he Web o Science and Scopus da abases. A o al o 78 pee - e iewed a icles we e
analyzed and ca ego ized s a es and ends in o 10 dimensions: (1) echnology eadiness, (2) cus omiza ion, (3)
AI ools and needs, (4) da a equi emen s, (5) skills and compe encies, (6) inancial eadiness, (7) managemen
suppo , (8) ma ke and compe i i e p essu e, (9) pa ne ship and collabo a ion, and (10) egula o y compliance,
based on he echnology–o ganiza ion–en i onmen (TOE) heo e ical model. A bibliome ic mapping app oach
was adop ed o isualize bibliome ic da a using VOS iewe . The e iew b ings oge he collec i e insigh s om
se e al leading expe con ibu o s o emphasize a eas whe e SMEs need addi ional suppo o ully le e age AI
echnologies. The esul s p o ide p agma ic insigh s o policymake s, helping hem de elop ailo ed app oaches
o bo h SMEs and la ge en e p ises o mee hei unique needs while acknowledging AI’s undeniable ole in
compe i i eness and g ow h.
In oduc ion
In he pas 10 yea s, s a e-o - he-a echnologies such as a i icial
in elligence (AI), da a analy ics, and machine lea ning ools ha e
e olu ionized he pe o mance o o ganiza ions om op o bo om
ac oss business unc ions (Hwang & Kim, 2021). The ex ensi e imple-
men a ion o such echnologies wi hin i ms gene a es highe e ec-
i eness, inc eases e iciency, and d i es o e all p oduc i i y (Cza ni zki
e al., 2023; Damioli e al., 2021). Owing o he complexi y o business,
da a a ailabili y, sophis ica ed echniques, and in as uc u e ad ance-
men , AI has quickly eme ged as a op echnological ocus o socie y
and o ganiza ions o s eamline ope a ions, make da a-d i en decisions,
and o e pe sonalized solu ions a scale (Hwang & Kim, 2021; Kuma
e al., 2024; Mak idakis, 2017; Mikale & Gup a, 2021).
To inc ease he alue o hei p oduc s and se ices, se e al com-
panies ha e made la ge in es men s in AI echnologies, making i a
c ucial componen o hei ope a ions (B ˘
a ucu e al., 2024). Nume ous
s udies ha e e ealed ha se e al in e nal and/o ex e nal ac o s in-
luence he use o AI in business ope a ions, including he o ganiza ional
se ing in which he echnology is used, he echnology i sel , and
en i onmen al aspec s (Baabdullah e al., 2021; Kulka ni e al., 2024;
Rana e al., 2024). Simila ly, esea ch documen s a b oad ange o i m
pe o mance ou pu s, such as inancial pe o mance (Ab okwah-La bi &
Awuku-La bi, 2024; Mousa e al., 2024), p o i abili y and cos s uc u e
(Wamba-Taguimdje e al., 2020), lea ning and inno a ion-enhanced
alues (Feng e al., 2024), and economic and ope a ional pe o mance
(Badghish & Soom o, 2024; Chen e al., 2024).
While AI has consis en ly been linked o posi i e o ganiza ional
ou comes, p o iding a s a egic ad an age o i ms o all sizes, he
dynamics o AI implemen a ion and i s ou comes can di e signi ican ly
be ween small and medium en e p ises (SMEs) and la ge , well-
es ablished i ms. These di e ences esul om a ia ions in e-
sou ces, expe ise, cos s uc u e, and suppo sys ems, among o he
ac o s (Ab okwah-La bi & Awuku-La bi, 2024; Cza ni zki e al., 2023;
* Co esponding au ho .
E-mail add ess: [email p o ec ed].
Con en s lis s a ailable a ScienceDi ec
Jou nal o Inno a ion & Knowledge
jou nal homepage: www.else ie .com/loca e/jik
h ps://doi.o g/10.1016/j.jik.2025.100682
Recei ed 14 No embe 2024; Accep ed 25 Feb ua y 2025
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
A ailable online 22 Ma ch 2025
2444-569X/© 2025 The Au ho (s). Published by Else ie España, S.L.U. on behal o Jou nal o Inno a ion & Knowledge. This is an open access a icle unde he
CC BY-NC-ND license (
h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/ ).
Damioli e al., 2021; Kopka & Fo nahl, 2024; Ramme e al., 2022;
Schwaeke e al., 2024; Wamba-Taguimdje e al., 2020). Fo ins ance,
Schwaeke e al. (2024) examined he cu en s a e o AI adop ion in
SMEs h ough a sys ema ic e iew and ound ha in as uc u e, cul-
u e, compa ibili y, and egula ions a e key ac o s in luencing AI
adop ion. Howe e , impo an componen s such as AI ools and needs,
da a equi emen s, and managemen suppo we e no included in hei
s udy. Inclusion o hese ac o s would ha e added u he dep h o he
unde s anding o AI adop ion pa e ns in SMEs. S udies also sugges ha
ac o s such as digi al cul u e, pa ne p essu e, and adop ion cos s play
a c ucial ole in SMEs’ adop ion o eme ging echnologies (Faiz e al.,
2024).
The con ibu ions o his sys ema ic e iew a e h ee old. Fi s ,
exis ing esea ch ou comes o en ocus exclusi ely on ei he SMEs (e.g.,
Baabdullah e al., 2021; C ocke e al., 2021; Lemos e al., 2022; Pe -
e z-Ande sson e al., 2024; Rawashdeh e al., 2023; Sha ma e al., 2022;
Wei & Pa do, 2022) o la ge i ms (e.g., Ahmad 2024; Bansal e al.
2024; Damioli e al. 2021; Fa es e al. 2023; Manse Payne e al. 2021;
Omoge e al. 2022; and Rahman e al. 2023), limi ing comp ehensi e
compa a i e insigh s ac oss o ganiza ional sizes. This dicho omy
p omp s impo an ques ions abou whe he AI ools and applica ion
modali ies a e inhe en ly simila o whe he signi ican a ia ions in
implemen a ion and ou comes s em om di e ing o ganiza ional dy-
namics. Second, al hough p e ious s udies ecognize AI’s ole in
enhancing ope a ional and s a egic pe o mance, li le esea ch con-
solida es he speci ic o ganiza ional, echnological, and en i onmen al
ac o s ha in luence AI adop ion ac oss a ious i m sizes. Signi ican
knowledge gaps emain ega ding how SMEs and la ge i ms can ailo
hei s a egies o he bes possible AI in eg a ion h ough compa able
amewo ks.
Thi d, while gene al academic knowledge is g owing apidly, AI-
ela ed esea ch in he business ield is occu ing mo e apidly. B eak-
h oughs in AI- ela ed a eas occu much mo e equen ly han in o he
ields. Wi h so much new in o ma ion being p oduced, a sys ema ic
li e a u e e iew is no only ele an bu also necessa y o dis ill and
comp ehend he mos c i ical indings (Linnenluecke e al., 2020). I
allows o a comp ehensi e analysis o many ecen s udies, syn hesizing
he mos c i ical indings in o a cohe en na a i e a he han dupli-
ca ing e o s. This e iew b ings oge he collec i e insigh s om
se e al leading expe con ibu ions o discuss how AI is commonly
u ilized and how i should be cus omized o mo e speci ic needs in
a ious o ganiza ional se ings.
This s udy add esses hese gaps by sys ema ically e iewing and
ca ego izing 78 pee - e iewed a icles h ough he lens o he TOE
amewo k. A bibliome ic mapping app oach was used o isualize
bibliome ic da a and he esul s o a sys ema ic li e a u e e iew.
Robus compa isons by in eg a ing bibliome ic analysis wi h a hema ic
e iew con ibu e o a nuanced unde s anding o AI adop ion dynamics,
he eby helping o iden i y c i ical enable s and ba ie s as well as
simila i ies and di e ences in he adop ion o AI o each ype o en-
e p ise. Unde s anding such dynamics is e y impo an in de ising
s a egies ha could enhance he smoo h use o AI ac oss a ied o ga-
niza ional sizes and ensu e ha small businesses a e also compe i i e in
an inc easingly AI-d i en business.
The emaining sec ions o he s udy a e s uc u ed as ollows. In
Sec ion 2, key heo e ical deba es and o e all hemes wi hin he li e a-
u e a e p esen ed. Sec ion 3 de ails he ma e ials and me hods used in
di e en s ages o he e iew p ocess. Sec ion 4 p esen s a discussion o
he key clus e s ela ed o AI ools, applica ions, and modali ies. The
s udy concludes in Sec ion 5.
Resea ch ques ions
In ligh o he abo e discussion, he pape add essed he ollowing
esea ch ques ions by ho oughly examining 78 pee - e iewed s udies
and ca ego izing s a es and ends in o 10 dimensions h ough he lens
o he TOE amewo k:
1. Wha a e he key enable s and de e en s o AI adop ion in SMEs and
how do hey compa e o hose o la ge , well-es ablished en e p ises?
2. A e he ac o s in luencing he adop ion o AI ools by SMEs and la ge
i ms simila o signi ican ly di e en ?
Li e a u e e iew
De ini ion, scope, and cha ac e is ics o AI
AI is an in o ma ion communica ion echnology ha can pe o m
asks independen ly and no mally equi es human in elligence o make
decisions, c ea e g ea e e iciencies, and enhance p oduc i i y
(A akpogun e al., 2021; Ghosh e al., 2018). O he s ha e de ined i as
he capaci y o a machine o hink like and imi a e human in elligence
(Va ma e al., 2024). Haenlein and Kaplan (2019) asse ha i is he
sys em’s capaci y o accu a ely e alua e da a and lea n om i o
accomplish objec i es.
F om au oma ing epe i i e ope a ions o enhancing human abili ies
in complica ed se ings including image iden i ica ion, p ocessing,
decision-making, na u al language p ocessing, and speech syn hesis, AI
spans a wide ange o sec o s and applica ions. In business, he po en ial
impac is as , in luencing unc ions such as ma ke ing (Ab okwah-La bi
& Awuku-La bi, 2024; Kuma e al., 2024), p oduc ion (Cha e jee e al.,
2021c; Me hi & Ha ouche, 2024), human esou ces (Li e al., 2023;
Kapoo , 2024; Vedap adha e al., 2024), secu i y (Rawinda an e al.,
2022), and inno a ion ac i i ies and beyond (Feng e al., 2024; Ramme
e al., 2022).
AI adop ion up ake in con empo a y business
AI co e s a a ie y o indus ies and applica ions, such as image
ecogni ion, p ocessing, decision-making, na u al language gene a ion,
and speech syn hesis, anging om au oma ing epe i i e asks o aug-
men ing human capabili ies in di e en domains like business, heal h-
ca e, enginee ing, and echnology (Raman e al., 2024). AI in es men s
and comme cial applica ions ha e su ged d ama ically ac oss se e al
indus ies o e he las 10 yea s (Babina e al., 2024). Acco ding o he
li e a u e, i ms epo g ea e g ow h in he use o AI echnologies in
di e en business ope a ions and unc ions, wi h annual e enue g ow h
inc easing by 30% mo e o AI adop e s han non-AI adop e s (Lee e al.,
2022). The epo u he unde sco ed he cos sa ings and e enue
g ow h in businesses whe e AI was used. The ph ase AI adop ion de-
sc ibes he ini ial s age in which a i m in eg a es AI in o i s business
p ocess. The e m u iliza ion e e s o he ac ual implemen a ion o AI
ools in he i m’s daily ac i i ies which in ol es he applica ion, in e-
g a ion, and usage o AI in exis ing p ocesses and sys ems (Tominc e al.,
2024).
Technology-o ganiza ion-en i onmen (TOE) amewo k
Va ious amewo ks, especially he echnology accep ance model
(TAM) de eloped by Da is e al. (1989) and i s ex ensions, ha e in es-
iga ed he d i e s o use s’ accep ance and adop ion o eme ging
echnologies (Venka esh & Bala, 2008; Venka esh e al., 2012). Ven-
ka esh e al. (2003) also examined echnology adop ion a he indi idual
le el using uni ied heo y o adop ion and use o echnology (UTAUT).
O he models, such as he di usion o inno a ions (DOI), emphasize
mo e aspec s ela ed o he sp ead o inno a ions wi hin social sys ems
(A ko ul e al., 2021). Taken oge he , hese heo ies p o ide e y
powe ul ools o unde s anding use beha io , bu o en along speci ic
dimensions o echnology adop ion.
Simila ly, he TOE model was de eloped o explain he in luence o a
b oad composi e o ac o s on he adop ion and p ocesses o echno-
logical inno a ion: (1) hose ela ed o he cha ac e is ics o echnology
i sel , (2) o ganiza ional con ex s, and (3) he ex e nal en i onmen
wi hin which an o ganiza ion ope a es (Bake , 2012). The amewo k
has al eady been es ed empi ically o alidi y and eliabili y in se e al
S.G. Ayinaddis
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
2
p e ious s udies and ecognized as one o he s onges heo e ical ools
o explain echnology implemen a ion beha io in i ms (Cha e jee
e al., 2021c; Das & Bala, 2024; Ganguly, 2024; Mish a & Pa hak, 2024;
Nguyen e al., 2022).
Despi e i s s ong heo e ical basis o analyze echnological adop ion,
he TOE amewo k has ce ain limi a ions. The TOE amewo k is said
o o e simpli y he complex iadic ela ionship cons ella ions o he
h ee elemen s in ol ed ( echnology, o ganiza ion, and en i onmen ),
assuming ha hey a e igid and well-dis inguished segmen s. Acco ding
o Gwaka e al. (2023), he o he limi a ion o his amewo k is a limi ed
iew on socioeconomic and cul u al ac o s ha can in luence he
echnology adop ion p ocess. In he same ein, o he indings c i icize
i s eliance on quan i a i e iews ha may blindside he quali a i e
aspec s o he echnology adop ion p ocess (Bake , 2012; Lin & Chen,
2023).
In o de o concep ualize he di e ences in AI adop ion be ween
SMEs and la ge en e p ises, as shown in Fig. 1, he pape build upon he
TOE heo e ical model o his s udy.
Ma e ials and me hods
The PRISMA amewo k– he p e e ed epo ing me hod o sys-
ema ic e iews and me a-analysis, was used o ensu e ha ou me hods
we e me hodologically sound and enabled acking da a p og ess a
di e en s ages o he e iew p ocess (Mohe e al., 2015). Addi ionally,
he bibliome ic mapping app oach was used o ep esen bibliome ic
da a and esul s o a sys ema ic li e a u e e iew using VOS iewe
so wa e.
By analyzing and in e p e ing p io esea ch in a pa icula domain,
his me hod p o ides (1) epea able, (2) ep oducible (K aus e al.,
2023), and (3) eliable esul s (Snyde , 2019). I is conside ed he mos
igo ous echnique because i p o ides eliable, unbiased, and ep o-
ducible esul s, hus allowing a comp ehensi e syn hesis o he exis ing
e idence o add ess he esea ch objec i e (T an ield e al., 2003). The
e iew p ocesse began by de ining he esea ch p oblem and objec i e.
Second, a li e a u e sea ch was conduc ed based on PRISMA p inciples.
Thi d, he da a was syn hesized and analyzed. Las ly, meaning ul
discussion and conclusions we e p o ided, as shown in Fig. 2.
Da a ex ac ion and app aisal
In a sys ema ic li e a u e e iew app oach, he e a e wo impo an
elemen s: (1) se ing c i e ia o he inclusion and exclusion o eco ds,
and (2) e alua ion o he quali y o he selec ed eco ds (Linnenluecke
e al., 2020; Mohe e al., 2015; Snyde , 2019). The da abases chosen
we e Scopus and Web o Science (WOS), so he a ailable in o ma ion
may complemen each o he o e ie e high-quali y jou nals. These
da abases con ain op schola ly jou nals and p o ide s able co e age,
making hem sui able o in-dep h analysis (Ga g e al., 2024; Ha zing &
Alakangas, 2016; Raman e al., 2024).
Fi s , a da abase sea ch was conduc ed o h ee main domain a eas
wi h Boolean sea ch e ms in he i le, abs ac , and keywo ds sec ions o
co e all aspec s comp ehensi ely: (1) di e en men ions o AI such as
a i icial in elligence, AI, machine lea ning, and deep lea ning, as
applied by he p io li e a u e (Schwaeke e al., 2024); (2) p ocesses o
ac ions ega ding he e ms o AI adop ion such as adop ion, imple-
men a ion, usage, and u iliza ion, based on he p e ious s udies; and (3)
i m size- ela ed e ms such as small and medium en e p ises, small and
medium-sized en e p ises, SMEs, SME, small and medium businesses,
la ge en e p ises, la ge i ms, la ge businesses, mul ina ional companies,
global companies, and la ge o ganiza ions. These e ms and ph ases a e
ele an o he esea ch objec i e o unde s anding how AI adop ion
dynamics may di e ac oss i m sizes (see he ull syn ax used o access
eco ds in Appendix A).
In he nex s ep, h ee eligibili y c i e ia we e es ablished o he i s
sea ch: (1) only a icles w i en in English, (2) s udies conduc ed in he
p e ious 10 yea s (2015–2024), and (3) in he subjec a ea o business,
managemen , accoun ing, and compu e science. Only a icles classi ied
as jou nal documen s we e included, excluding e iew a icles, le e s o
he edi o , commen a ies, g ay li e a u e, case epo s, and duplica es.
The sea ch esul we e u he e ined by elimina ing keywo ds ha
we e un ela ed o he opic. Finally, by eading he i les and abs ac s o
he emaining eco ds, a inal sample o 78 pee - e iewed a icles was
selec ed (Fig. 3).
Fig. 1. Concep ual model o he s udy.
Sou ce: The concep ual model is based on he TOE amewo k adap ed om Nguyen e al. (2022).
S.G. Ayinaddis
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
3
The s udy was limi ed o pape s published a e 2015 because o he
quick de elopmen o AI echnologies and hei subs an ial in luence on
business a his pe iod (Mai e al., 2024, Babina e al., 2024). O e he
pas 10 yea s, businesses ha e inc easingly used AI echnologies such as
machine lea ning, deep lea ning, and p edic i e analy ics (Shao e al.,
2022). This pe iod has also wi nessed a ise in esea ch in o he use o AI
by businesses o all sizes, om s a ups o global conglome a es, in
addi ion o he de elopmen o egula ions and e hical s anda ds o AI
in business (Tominc e al., 2024).
Da a cleaning and coding
Gi en he goal o ou e iew and he need o p esen esul s based on
a sys ema ic and unbiased analysis o he li e a u e (T an ield e al.,
2003), in he hi d s ep, each o he 78 pape s was assigned a unique ID
numbe a e ex ac ion o acili a e smoo h iden i ica ion among he
a icles. ID numbe s emained cons an h oughou he e iew p ocess.
The ollowing alida ion columns we e hen used o a ange he da a: ID
numbe , au ho s, i le, publica ion yea , keywo ds, abs ac , and jou nal
name.
Resul s and discussion
Dis ibu ion o s udies based on le el o analysis
Fig. 4 illus a es he p opo ion o AI adop ion s udies ac oss
di e en i m sizes, ca ego ized as SMEs, la ge i ms, bo h (SMEs and
la ge), and uniden i ied. Based on hese esul s, we can de e mine which
i m sizes a e mos ep esen ed among he 78 eco ds analyzed in he
cu en sys ema ic e iew. Acco dingly, a signi ican po ion o AI
adop ion s udies ocuses on he SMEs ca ego y (44%), indica ing a
g owing in e es in unde s anding how AI is being u ilized wi hin hese
i ms. This end is likely d i en by he unique challenges SMEs ace
compa ed wi h la ge en e p ises. S udies ha do no speci y a pa icula
i m size (labelled as no iden i ied) accoun o 32% o he o al. La ge
i ms cons i u e 15% o he eco ds, and s udies examining bo h SMEs
and la ge en e p ises ep esen 9%. The e o e, he dis ibu ion o hese
s udies ac oss di e en i m sizes is ele an o he cu en sys ema ic
li e a u e e iew objec i e o unde s anding he dynamics o AI adop-
ion in SMEs and la ge i ms (see ull lis eco ds wi h he co esponding
le el o analysis in Appendix B).
Yea o publica ion o selec ed s udies
No publica ions we e eco ded om he selec ed s udies in his e-
iew, no publica ions we e eco ded om 2015 o 2018. Howe e , a
signi ican accele a ion in esea ch ac i i ies o e he las ew yea s has
been obse ed, peaking in 2024 (Fig. 5).
Publica ion dis ibu ion based on he heo ies used
The heo e ical models employed by he au ho s in he selec ed e-
co ds we e e iewed o ob ain a ho ough unde s anding o he unde -
lying heo e ical models used in he s udies. As indica ed in Fig 6, hose
s udies coun ed as no applicable we e hose in which he au ho s did no
explici ly indica e he heo ies used in hei s udy. Howe e , he TOE
amewo k is he mos commonly used heo e ical model, ollowed by
he echnology accep ance model (TAM).
Bibliome ic analysis
A bibliome ic analysis, a popula quan i a i e echnique in sys-
ema ic li e a u e e iew esea ch, was conduc ed. P e ious s udies,
such as Linnenluecke e al. (2020), sugges ha his me hod is e ec i e
in mapping he heme o in e es o see i s in ellec ual oo s and he
s uc u e o he li e a u e o e ime, helping esea che s iden i y key
hemes and ele an keywo ds o analysis. The bibliome ic app oach
also acili a es ne wo k analysis and isualiza ion.
Keywo ds ne wo k analysis
A keywo d co-occu ence analysis was conduc ed using a ull-
coun ing app oach o explo e he mos p e alen hemes and key-
wo ds in AI adop ion. To a oid bias du ing he keywo d selec ion p o-
cess, no only he abs ac s and keywo ds o he selec ed eco ds bu also
he en i e ex we e ca e ully examined. Fig. 7 p esen s he mos
equen ly used keywo ds om bo h he WOS and Scopus in he s udy
om 2015 o 2024. Among he keywo ds, a i icial in elligence is he mos
men ioned, wi h 51 occu ences, ollowed by echnology adop ion and
SMEs, wi h 14 occu ences each.
Fig. 2. The e iew p ocesses.
S.G. Ayinaddis
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
4
Analysis o he clus e s acco ding o TOE model
Technological ac o s
Technology eadiness. Technology eadiness is di ec ly linked o IT
in as uc u e. Se e al s udies ha e emphasized he impo ance o he
digi al ma u i y le el (B ˘
a ucu e al., 2024) and ad anced digi al
in as uc u e o enable echnological eadiness in he adop ion o AI
ac oss a ious sec o s (Aga wal, 2022; Baabdullah e al., 2021; Das &
Bala, 2024; Do a e al., 2022; Issa e al., 2022; Ko iˇ
c e al., 2024; Me hi
& Ha ouche, 2024; Tominc e al., 2024). SMEs and la ge i ms ha e
di e en in as uc u e equi emen s o success ully implemen ing AI
(Badghish & Soom o, 2024). La ge en e p ises ypically possess he
necessa y echnological in as uc u e o suppo ex ensi e AI sys ems.
Aghimien e al. (2024), Gup a e al. (2022), Man i & Mish a (2023),
Solaimani & Swaak (2023), and Tominc e al. (2024) ag ee ha he
subs an ial inancial esou ces a ailable o la ge i ms enable hem o
le e age ad anced AI echnologies ha equi e signi ican compu a-
ional powe and da a o aining and deploymen . This capabili y al-
lows la ge o ganiza ions o imbibe AI in o hei sys ems, aising
e iciency and inno a ion.
SMEs usually ace high ba ie s o accessing he necessa y in a-
s uc u al acili ies o AI adop ion (Kapoo , 2024; Schlegel e al., 2023;
Tawil e al., 2024). Acco ding o Jalil e al. (2024), AI eadiness is ound
complemen a y o echnological o ien a ion in SMEs, among o he ac-
o s (Polise y e al., 2024).
Fig. 3. S udy selec ion p ocess (PRISMA low diag am).
S.G. Ayinaddis
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
5
Sys em cus omiza ion ( lexibili y)
Many SMEs see ease o use as a key c i e ion in choosing AI ools o
simpli y hei adop ion (Hamdan e al., 2022b; Vedap adha e al., 2024).
The eason is ha SMEs need AI ools ha equi e li le aining o
specialized skills gi en hei limi ed alen esou ces. Sha ma e al.
(2022) ound ha he bigges enable o de e en o SMEs’ in en ion o
adop AI is he sys em cha ac e is ics o echnical con ex , u he sup-
po ed by Ba a a e al. (2023), Handoko (2021), and Ma ou khani e al.
(2023). Hansen and Bøgh (2021) no ed ha he eason behind he mos
success ul implemen a ion o AI in SMEs is i s g ea e simplici y. They
usually choose plug-and-play AI ools, such as cha bo s (Sha ma e al.,
2022) and hey will become hesi an i hey belie e ha he sys em is
complica ed and challenging o use and apply. Use - iendly AI ools
wi h less complex in e aces allow SMEs o adop AI h ough basic
aining (Cha e jee e al., 2022; Cha e jee e al., 2021b; Hamdan e al.,
2022a; Ho e al., 2022). The e o e, he deg ee o lexibili y and
compa ibili y wi h exis ing sys ems a e s ong p edic o s o AI adop ion
in SMEs (Kaymakci e al., 2022; Rawashdeh e al., 2023).
AI ools and needs
Looking a he co e objec i es and scope o AI adop ion be ween
SMEs and la ge i ms is ano he impo an ac o . Analyzing he a io-
nale and pu pose o AI adop ion in e ms o he co e goals and ex en o
co e age o SMEs and la ge i ms is ano he ele an conside a ion. As
epo ed in p io s udies, he e a e s a is ically signi ican di e ences in
he AI needs/usage equi emen s be ween la ge i ms in e ms o i m
size and SMEs (Tominc e al., 2024). La ge en e p ises mainly imple-
men AI when handling big da a, in e ac ing wi h cus ome s om
a ound he wo ld and implemen ing logis ics on a la ge scale. These
goals equen ly ela e o managing a o dances o ope a ion and
op imizing he e iciency ha a ises om highe deg ees and b ead hs o
applica ion (Yang e al., 2024).
SMEs, despi e acknowledging AI’s bene i s, p io i ize immedia e
p ac ical conce ns (Tominc e al., 2024). Fo ins ance, hey equi e AI
o speci ic pu poses and conside ime as a cos (Rawashdeh e al.,
2023), such as cha bo s o enhance cus ome sa is ac ion, manage in-
en o y, o au oma e cle ical asks. Thei goals a e equen ly mo e
pa icula and empo a y owing o he ma ke ing a o dances ha in-
luence less dep h and wid h o use (Yang e al., 2024).
Fig. 4. P opo ion o he s udies based on i m size analysis.
NB: No iden i ied e e s o hose s udies in which he au ho s did no explici ly speci y he pa icula i m size such as small o la ge i ms in he pape .
Fig. 5. Dis ibu ion o publica ion o e 2015–2024.
Sou ce: Compiled by au ho s, 2024
S.G. Ayinaddis
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
6
Da a equi emen s
AI in eg a ion calls o he accumula ion o huge da a se s ha can be
used o un p edic i e analy ics (Fu e al., 2023). Big i ms o en accu-
mula e subs an ial da a, which helps pu AI’s da a p ocessing and
analy ical p ospec s o be e use han SMEs. This access o la ge da a
olumes esul s in imp o ed decision-making op ions and
be e -in o med decisions (Tominc e al., 2024). AI can also be
cus omized acco ding o he needs o la ge en e p ises as well as he way
hese companies wo k o mee hei needs, hus inc easing i s alue.
In con as , Pe e z-Ande sson e al. (2024) ound ha SMEs
encoun e di icul ies in ob aining and managing he as olumes o
da a needed o AI applica ions. Simila ly, p oblems wi h access o a big
da a se o AI deploymen in manu ac u ing i ms ha e been b ough o
ligh (Ko iˇ
c e al., 2024), and SMEs may encoun e da a a ailabili y
conce ns in some si ua ions.
O ganiza ional con ex s
Skills and compe encies
Skills and compe encies play an impo an ole in AI echnology
adop ion. Func ional and ope a ional knowledge c ea es di e ences
be ween SMEs and la ge i ms (G asho & Kopka, 2023; Wei & Pa do,
2022). Fo ins ance, Huseyn e al. (2024) highligh ed he posi i e ou -
comes o aining p og ams aimed a upg ading ac o s’ skill o e ec-
i ely u ilize AI in SME ope a ions. S udies ha e consis en ly shown ha
angible esou ces and wo k o ce skills posi i ely in luence AI imple-
men a ion (Chen e al., 2024). Knowledge embodimen a ec s people’s
in en ions o adop echnology (Pee e al., 2019). In ac , small i ms
o en lack such expe ise (Hansen & Bøgh, 2021), as hey migh no be
able o p o ide compe i i e sala ies and ca ee g ow h as la ge i ms do,
which can hinde hei abili y o implemen AI solu ions e ec i ely
(Pe e z-Ande sson e al., 2024). Reliance on ou side consul an s o
pa ne ships may esul om his lack o echnical expe ise, which is
no always p ac ical o smalle businesses because o hei inancial
cons ain s.
Con e sely, la ge i ms ypically ind i easie o inco po a e AI in o
hei business, as hey ha e he inancial means o e ain specialized
pe sonnel (Tominc e al., 2024). Bloms e and Koi um¨
aki (2022)
concep ualized he skills and compe encies equi ed o success ul AI
adop ion. As a esul , pe sonnel compe encies in he p ope ies o he
da a and hei abili y o manage success ul machine lea ning p ojec s.
The e o e, hi ing quali ied da a scien is s o enhance employees’ IT
awa eness and he igh se o skills posi i ely in luences hei a i udes
owa d AI adop ion (Almashaw eh e al., 2024; Khaliq e al., 2022;
Rahman e al., 2023; Solaimani & Swaak, 2023). Tawil e al. (2024)
sugges ed ha SMEs need he igh skills o p oduce use ul insigh s om
da a-d i en decision-making using AI.
Resou ces and inancial eadiness
Resou ces a e he ne e cen e s o e e y o ganiza ion. Resea ch in
inance has shown a s ong co ela ion be ween i m size and inancial
a ailabili y. In eg a ing AI equi es bo h in e nal and ex e nal in-
es men s, and inancial esou ces a e c i ical de e minan s. Schola s
ha e no ed ha he capi al elemen plays a c i ical ole in he alue-
enhancemen mechanism o AI ools (Luo & Yu, 2022). Acco ding o
Tominc e al. (2024), la ge i ms ha e he unds and human capi al o
in es in sophis ica ed AI echnologies. They may also de elop in-house
AI ools adap ed o hei needs.
In mos cases, such in es men s a e beyond he scope o SMEs
because o hei limi ed esou ces and es ic ed access o inancing (Bąk
e al., 2024; Tominc e al., 2024). They a e hesi an o in es in so wa e
and ha dwa e i hey canno expec quick posi i e esul s and e enue
(Tawil e al., 2024). Budge cons ain s (Wong e al., 2020) and he
pe cei ed cos o AI (Mousa e al., 2024; Sha ma e al., 2022) a e acu e
o SMEs, and hey a ely ha e in-house AI solu ions, leading hem o
ind mo e a o dable, basic AI solu ions. Due o hei inhe en s uc u e,
SMEs ace di icul ies in accessing inancial esou ces as hey lack
colla e al o secu e loans (Nagy e al., 2023; Wang & Pan, 2022). A s udy
Fig. 6. Publica ion dis ibu ion based on he heo ies used.
Sou ce: Compiled by au ho s, 2024
S.G. Ayinaddis
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
7
Fig. 7. Co-occu ence o keywo ds based on ull-coun ing me hod.
Sou ce: VOS iewe s’ esul compiled by au ho s, 2024
S.G. Ayinaddis
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
8
In e na ional Jou nal o Da a and Ne wo k Science, 7(1), 25–34. h ps://doi.o g/
10.5267/j.ijdns.2022.12.010
Rawinda an, N., Jayal, A., & P akash, E. (2022). Explo a ion o he impac o
cybe secu i y awa eness on small and medium en e p ises (SMEs) in Wales using
in elligen so wa e o comba cybe c ime. Compu e s, 11(12), 174. h ps://doi.o g/
10.3390/compu e s11120174
Schlegel, D., Schule , K., & Wes enbe ge , J. (2023). Failu e ac o s o AI p ojec s: esul s
om expe in e iews. In e na ional Jou nal O In o ma ion Sys ems And P ojec
Managemen : IJISPM, 11(3), 25–40. h ps://doi.o g/10.12821/ijispm110302
Schwaeke, J., Pe e s, A., Kanbach, D. K., K aus, S., & Jones, P. (2024). The new no mal:
The s a us quo o AI adop ion in SMEs. Jou nal o small business managemen , 1–35.
h ps://doi.o g/10.1080/00472778.2024.2379999
Shao, Z., Zhao, R., Yuan, S., Ding, M., & Wang, Y. (2022). T acing he e olu ion o AI in
he pas decade and o ecas ing he eme ging ends. Expe Sys ems wi h Applica ions,
209, A icle 118221. h ps://doi.o g/10.1016/j.eswa.2022.118221
Sha ma, S., Singh, G., Islam, N., & Dhi , A. (2022). Why do SMEs adop a i icial
in elligence-based cha bo s? IEEE T ansac ions on Enginee ing Managemen , 71,
1773–1786. h ps://doi.o g/10.1109/TEM.2022.3203469
Singh, A., Dwi edi, A., Ag awal, D., & Singh, D. (2023). Iden i ying issues in adop ion o
AI p ac ices in cons uc ion supply chains: owa ds managing sus ainabili y.
Ope a ions Managemen Resea ch, 16(4), 1667–1683. h ps://doi.o g/10.1007/
s12063-022-00344-x
Singh, N., Jain, M., Kamal, M. M., Bodhi, R., & Gup a, B. (2024). Technological
pa adoxes and a i icial in elligence implemen a ion in heal hca e. An applica ion o
pa adox heo y. Technological Fo ecas ing and Social Change, 198, A icle 122967.
h ps://doi.o g/10.1016/j. ech o e.2023.122967
Snyde , H. (2019). Li e a u e e iew as a esea ch me hodology: An o e iew and
guidelines. Jou nal o business esea ch, 104, 333–339. h ps://doi.o g/10.1016/j.
jbus es.2019.07.039
Solaimani, S., & Swaak, L. (2023). C i ical success ac o s in a mul i-s age adop ion o
a i icial in elligence: A necessa y condi ion analysis. Jou nal o Enginee ing and
Technology Managemen , 69, A icle 101760. h ps://doi.o g/10.1016/j.
jeng ecman.2023.101760
Tawil, A.-R. H., Mohamed, M., Schmoo , X., Vlachos, K., & Haida , D. (2024). T ends and
challenges owa ds e ec i e da a-d i en decision making in UK Small and Medium-
sized En e p ises: Case s udies and lessons lea n om he analysis o 85 Small and
Medium-sized En e p ises. Big Da a and Cogni i e Compu ing, 8(7), 79. h ps://doi.
o g/10.3390/bdcc8070079
Tominc, P., O eˇ
ski, D., ˇ
Canˇ
ce , V., & Roˇ
zman, M. (2024). S a is ically signi ican
di e ences in AI suppo le els o p ojec managemen be ween SMEs and la ge
en e p ises. AI, 5(1), 136–157. h ps://doi.o g/10.3390/ai5010008
T an ield, D., Denye , D., & Sma , P. (2003). Towa ds a me hodology o de eloping
e idence-in o med managemen knowledge by means o sys ema ic e iew. B i ish
Jou nal o Managemen , 14(3), 207–222. h ps://doi.o g/10.1111/1467-8551.00375
Va ma, A., Pe ei a, V., & Pa el, P. (2024). A i icial in elligence and pe o mance
managemen . O ganiza ional Dynamics, 53, A icle 101037. h ps://doi.o g/10.1016/
j.o gdyn.2024.101037
Vedap adha, R., Ha iha an, R., Sudha, E., & Di yash ee, V. (2024). A i icial
in elligence–Talen acquisi ion in HEIs ec ui men s. The In e na ional Jou nal o
In o ma ion and Lea ning Technology(ahead-o -p in ). h ps://doi.o g/10.1108/IJILT-
09-2023-0176
Venka esh, V., & Bala, H. (2008). Technology accep ance model 3 and a esea ch agenda
on in e en ions. Decision Sciences, 39(2), 273–315. h ps://doi.o g/10.1111/j.1540-
5915.2008.00192.x
Venka esh, V., Mo is, M. G., Da is, G. B., & Da is, F. D. (2003). Use accep ance o
in o ma ion echnology: Towa d a uni ied iew. MIS qua e ly, 425–478. h ps://doi.
o g/10.2307/30036540
Venka esh, V., Thong, J. Y., & Xu, X. (2012). Consume accep ance and use o
in o ma ion echnology: Ex ending he uni ied heo y o accep ance and use o
echnology. MIS Qua e ly, 157–178. h ps://doi.o g/10.2307/41410412
Volkma , G., Fische , P. M., & Reinecke, S. (2022). A i icial in elligence and machine
lea ning: Explo ing d i e s, ba ie s, and u u e de elopmen s in ma ke ing
managemen . Jou nal o business esea ch, 149, 599–614. h ps://doi.o g/10.1016/j.
jbus es.2022.04.007
Wamba-Taguimdje, S.-L., Wamba, S. F., Kamdjoug, J. R. K., & Wanko, C. E. T. (2020).
In luence o a i icial in elligence (AI) on i m pe o mance: he business alue o AI-
based ans o ma ion p ojec s. Business p ocess managemen jou nal, 26(7),
1893–1924. h ps://doi.o g/10.1108/BPMJ-10-2019-0411
Wang, J. (2024). Using a i icial in elligence o analyze SME e-comme ce u iliza ion and
g ow h s a egies. Jou nal o Compu a ional Me hods in Sciences and Enginee ing, 24
(1), 611–621. h ps://doi.o g/10.3233/JCM-226933
Wang, M., & Pan, X. (2022). D i e s o a i icial in elligence and hei e ec s on supply
chain esilience and pe o mance: an empi ical analysis on an eme ging ma ke .
Sus ainabili y, 14(24), 16836. h ps://doi.o g/10.3390/su142416836
Wei, R., & Pa do, C. (2022). A i icial in elligence and SMEs: How can B2B SMEs le e age
AI pla o ms o in eg a e AI echnologies? Indus ial Ma ke ing Managemen , 107,
466–483. h ps://doi.o g/10.1016/j.indma man.2022.10.008
Wong, L.-W., Leong, L.-Y., Hew, J.-J., Tan, G. W.-H., & Ooi, K.-B. (2020). Time o seize
he digi al e olu ion: Adop ion o blockchain in ope a ions and supply chain
managemen among Malaysian SMEs. In e na ional Jou nal o In o ma ion
Managemen , 52, A icle 101997. h ps://doi.o g/10.1016/j.ijin omg .2019.08.005
Yang, J., Bloun , Y., & Am ollahi, A. (2024). A i icial in elligence adop ion in a
p o essional se ice indus y: A mul iple case s udy. Technological Fo ecas ing and
Social Change, 201, A icle 123251. h ps://doi.o g/10.1016/j.
ech o e.2024.123251
S.G. Ayinaddis
Jou nal o Inno a ion & Knowledge 10 (2025) 100682
15
