Jupić, Nedžad; Gadžo, Am a
Con e ence Pape — Accep ed Manusc ip (Pos p in )
Enhancing small and medium en e p ise pe o mance
h ough a i icial in elligence in eg a ion in accoun ing
Sugges ed Ci a ion: Jupić, Nedžad; Gadžo, Am a (2025) : Enhancing small and medium en e p ise
pe o mance h ough a i icial in elligence in eg a ion in accoun ing, Scien i ic con e ence
wi h in e na ional pa icipa ion SMEPP 2025: "A i icial in elligence as a D i ing Fo ce o he
De elopmen o Unde de eloped A eas: Oppo uni ies and Pe spec i es", No i Paza , Se bia, 3. jun.
2025, In: Uni e zi e u No om Paza u (Ed.): SMEPP 2005: zbo nik ado a, ISBN 978-86-83074-11-2,
Uni e zi e u No om Paza u, Depa man za ekonomske i ačuna ske nauke, No i Paza , S bija, pp.
33-46,
h ps://publikacije.uninp.edu. s/index.php/smepp2025/a icle/ iew/327
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/324135
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/4.0/
ENHANCING SMALL AND MEDIUM ENTERPRISE
PERFORMANCE THROUGH ARTIFICIAL INTELLIGENCE
INTEGRATION IN ACCOUNTING
Nedžad Jupić
Agency o Accoun ing and Business Consul ing Co ex LLC
Tuzla, Bosnia and He zego ina
nedzad[email p o ec ed]
Am a Gadžo
Facul y o Economics, Uni e si y o Tuzla
Tuzla, Bosnia and He zego ina
am a.ga[email p o ec ed]
Abs ac
This pape examines he deg ee o implemen a ion o a i icial in elligence (AI) ools wi hin
accoun ing in o ma ion sys ems in small and medium-sized en e p ises (SMEs) in Bosnia and
He zego ina, wi h he objec i e o iden i ying he pe cei ed bene i s, key ba ie s, and hei
e ec s on business pe o mance. The esea ch was conduc ed on a sample o 99 en e p ises,
u ilising a s uc u ed ques ionnai e. Findings e eal ha only 21.21% o en e p ises cu en ly
employ AI ools in hei accoun ing p ocesses, while a me e 20% o esponden s make use o
ad anced documen managemen sys ems (DMS) o cloud-based solu ions— echnologies ha
acili a e in eg a ion wi h AI. The esea ch highligh s a gene ally posi i e a i ude among
esponden s ega ding he impac o AI applica ion on SME ope a ions. The highes a e age
a ing was a ibu ed o he s a emen ha AI enhances he e iciency o accoun ing p ocesses
(4.22), ollowed by imp o ed quali y o inancial epo ing (4.02) and mo e e ec i e
manage ial decision-making (3.93), indica ing a ecognised added alue o AI ools in he a eas
o analy ics and decision suppo . The main obs acles o AI adop ion, as iden i ied by
esponden s, include a lack o knowledge and expe ise (3.80) and insu icien egula o y
amewo k (3.73). In e ms o inancial eadiness o in es in AI echnologies, 44.40% o
en e p ises indica ed willingness o in es up o app oxima ely EUR 500 annually, 37.40%
be ween EUR 500 and 1,000, and only 18.20% mo e han EUR 1,000.
O e all, he indings sugges a signi ican , ye unde u ilised po en ial o AI ools in SME
accoun ing, cha ac e ised by a ou able use pe cep ions bu cons ained by educa ional,
inancial, and legisla i e limi a ions.
Keywo ds: AI ools, accoun ing digi alisa ion, business enhancemen , small and
medium-sized en e p ises.
INTRODUCTION
The applica ion o a i icial in elligence (AI) in business p ocesses has signi ican ly
eshaped he way en e p ises ope a e. In he ield o accoun ing, AI ools enable he
au oma ion o ou ine asks, enhance he accu acy o inancial da a, and acili a e mo e
e icien and imely business decision-making. Fu he mo e, he digi al
ans o ma ion o ax au ho i ies and e o s o es ablish eal- ime ansac ion
moni o ing sys ems—p ima ily aimed a educing ax e asion—unde sco e he need
o he adop ion o mode n digi al solu ions. Digi alisa ion is, in ac , a p e equisi e
o he au oma ion o accoun ing p ocesses and he in eg a ion o AI ools in o
accoun ing in o ma ion sys ems. Howe e , small and medium-sized en e p ises
(SMEs) o en lag behind, i s ly in e ms o digi al ans o ma ion, and subsequen ly
in he adop ion o a i icial in elligence. Bosnia and He zego ina, as a ansi ional
economy, aces challenges in he digi al ans o ma ion o i s SME sec o . SMEs in
Bosnia and He zego ina a e cha ac e ised by limi ed esou ces, a sho age o quali ied
pe sonnel, and o en inadequa e suppo in e ms o in as uc u e and legal
amewo ks. Acco dingly, he aim o his pape is o analyse he ex en o which AI
ools a e applied in he accoun ing p ac ices o SMEs in Bosnia and He zego ina, and
o iden i y he pe cei ed bene i s and obs acles o hei implemen a ion. The pape is
s uc u ed in o h ee main sec ions. The i s sec ion p o ides an o e iew o p e ious
esea ch on he ad an ages and challenges o in oducing AI ools in accoun ing, along
wi h a e iew o s a is ical da a conce ning he le el o digi al ans o ma ion and he
use o AI ools by SMEs in he Wes e n Balkans. The second sec ion ou lines he
esea ch me hodology, objec i es, and da a sou ces. The hi d sec ion p esen s he
esea ch indings and o e s a discussion o he esul s.
LITERATURE REVIEW
A i icial in elligence (AI) signi ican ly enhances business ope a ions by enabling
mo e e icien da a managemen and he imp o emen o business p ocesses.
Howe e , he implemen a ion o AI also b ings challenges, including e hical conce ns
ela ed o da a p i acy and us (S anisic e al., 2024). Mo eo e , small and medium-
sized en e p ises (SMEs) o en lack he inancial and human esou ces necessa y o
adequa ely assess and apply he po en ial o AI applica ions o hei speci ic needs
(Szedlak e al., 2021). Acco ding o he WB DESI 2022 epo (p. 10), he adop ion o
digi al echnologies by SMEs in he Wes e n Balkans emains subs an ially below he
EU a e age (35% in he Wes e n Balkans compa ed o he EU a e age o 55%). On
a e age, only 7% o businesses in he Wes e n Balkans used big da a, 16% u ilised
cloud echnologies, and me ely 3% adop ed AI in 2021. In compa ison, hese igu es
in he EU we e 14%, 34%, and 8%, espec i ely. The same epo also s a es ha in
2021, 32% o SMEs in Bosnia and He zego ina had a leas a basic le el o digi al
in ensi y, closely aligning wi h he Wes e n Balkans a e age o 35%. Elec onic
in o ma ion exchange was used by 26% o en e p ises, which is sligh ly abo e he
egional a e age o 24%. Social media pla o ms we e no widely u ilised by
businesses in Bosnia and He zego ina, signi ican ly alling below he egional
a e age. Ad anced echnologies we e modes ly employed: 7% o businesses used
cloud solu ions, 4% big da a, and only 2% AI. Elec onic in oicing was used by 19%
o en e p ises, a highe a e han he egional a e age. Rega ding he use o
in o ma ion and communica ion echnologies (ICT) o en i onmen al sus ainabili y,
54% o en e p ises in Bosnia and He zego ina employed ICT o en i onmen al
p o ec ion pu poses, compa ed o he Wes e n Balkans a e age o 56% (WB DESI,
2022, p. 26).
Resea ch by Rawashdeha, Bakhi b, and Abaalkhailb (2022) iden i ied ou key
echnological ac o s—compa ibili y, challenge eadiness, e iciency imp o emen ,
and ime sa ing—as signi ican in in luencing he adop ion o AI by SMEs h ough
he au oma ion o accoun ing p ocesses. Accoun ing au oma ion no only acili a es
AI adop ion bu also independen ly inc eases in e es in he echnology. A i m's
eadiness o emb ace new echnologies and he alignmen o AI wi h exis ing business
p ac ices a e essen ial p e equisi es o success ul implemen a ion. Fu he mo e, ime
sa ings and imp o ed e iciency ha e a di ec impac on he decision o adop AI,
sugges ing ha au oma ion is no he sole de e minan . AI simpli ies p ocesses,
accele a es decision-making, and imp o es he accu acy o inancial epo ing
(Panwa , 2023). Acco ding o esea ch by Okeke, Baka e, and Achumie (2024), AI-
based accoun ing ools a e e olu ionising ou ine asks such as bookkeeping,
in oicing, and cos acking. By au oma ing hese unc ions, SMEs can educe human
e o s, sa e ime, and alloca e esou ces mo e e icien ly. This au oma ion enhances
e iciency, accu acy, and compe i i eness while educing s a wo kload (Khashimo
& Khayda o a, 2022). AI sys ems a e also capable o analysing la ge da ase s o de ec
anomalies and pa e ns in inancial da a, he eby imp o ing decision-making and
inancial sus ainabili y (Ga cía-Ve a, Juca-Maldonado & To es-Gallegos, 2023).
Iyelolu, Agu, Idemudia, and Ijomah (2024) highligh ha AI echnologies can
subs an ially enhance ope a ional e iciency, p oduc de elopmen , cus ome
engagemen , and compe i i e ad an age o SMEs.
RESEARCH OBJECTIVES, SAMPLE, AND DATA SOURCES
The subjec o his esea ch is he examina ion o he ex en o which a i icial
in elligence (AI) ools a e applied in he accoun ing p ac ices o small and medium-
sized en e p ises (SMEs), use s’ pe cep ions o hei use ulness, obs acles encoun e ed
du ing implemen a ion, and he analysis o he impac ha AI in eg a ion has no only
on he e iciency and quali y o accoun ing p ocesses bu also on he o e all business
pe o mance. The s udy se s ou he ollowing ope a ional esea ch objec i es:
• To de e mine he le el o applica ion o AI ools in accoun ing
• To assess use s’ pe cep ions o he con ibu ion o AI ools o SME accoun ing
• To iden i y ba ie s o he implemen a ion o AI ools in accoun ing
• To analyse he impac o AI ools on he o e all business pe o mance o SMEs
The esea ch popula ion consis ed o manage s and accoun an s wo king in SMEs in
Bosnia and He zego ina. Da a we e collec ed h ough an online su ey. The
ques ionnai e included a combina ion o Like scale ques ions, closed-ended
ques ions wi h mul iple-choice answe s, and open-ended ques ions designed o
p o ide quali a i e insigh s. The ques ionnai e was dis ibu ed o 550 email add esses,
and 99 esponses we e collec ed, esul ing in a esponse a e o 18%. The da a we e
analysed using desc ip i e s a is ics, co ela ion analysis, and eg ession analysis. To
p o ide a be e unde s anding o he esea ch esul s, he key cha ac e is ics o he
sample a e i s p esen ed. Among he su eyed en e p ises, 55.55% had been
ope a ing o mo e han 10 yea s; 17.17% had been in ope a ion be ween 5 and 10
yea s; and 27.28% had been ope a ing o up o 5 yea s. In e ms o sec o al
dis ibu ion, 66.67% o en e p ises ope a e in he se ice sec o , 15.15% in
manu ac u ing, and 18.18% in ade. The s uc u e o esponden s was as ollows:
36.36% accoun an s, 28.28% manage s, and 35.36% employees wo king in
accoun ing and inance depa men s.
RESEARCH FINDINGS AND DISCUSSION
In o de o comp ehensi ely explo e he ole o a i icial in elligence in SME
accoun ing p ac ices, his chap e ou lines he key ope a ional objec i es ha guide
he esea ch. These objec i es aim o p o ide s uc u ed insigh in o he ex en ,
pe cep ion, challenges, and e ec s o AI ool implemen a ion in he accoun ing
unc ion.
THE EXTENT OF APPLICATION OF AI TOOLS IN ACCOUNTING
When i comes o he use o a i icial in elligence (AI) ools—such as hose o ex ,
image, and code gene a ion (e.g. Cha GPT, Mic oso Copilo , Google Gemini),
de elope and da a analysis ools (e.g. Gi Hub Copilo , Da aRobo ), accoun ing and
inance ools (e.g. UiPa h, Xe o), and p edic i e analy ics pla o ms (e.g. Powe BI,
Sales o ce)— he indings indica e ha 55.56% o esponden s make use o a leas
one such ool, whe eas 44.44% do no . Howe e , he si ua ion appea s e en less
a ou able wi h espec o he applica ion o AI ools speci ically in accoun ing. In
esponse o he ques ion o whe he he en e p ise uses AI in accoun ing p ocesses—
such as op ical documen ecogni ion, au oma ic pos ing, p edic i e analy ics in
accoun ing, o cha bo suppo —only 21.21% o pa icipan s con i med usage, while
78.79% epo ed no applica ion o hese ools. Ne e heless, 16.16% o en e p ises
indica ed plans o in oduce AI in hei accoun ing unc ions, and 13.13% s a ed ha
hey cu en ly lack su icien knowledge on he opic. The emaining esponden s we e
undecided.
In o de o assess he p e equisi es o he in oduc ion o AI in accoun ing, a se ies o
ques ions was posed o explo e he cu en deg ee o digi alisa ion and au oma ion in
accoun ing p ocesses. One o he key inqui ies ela ed o he o ma in which
documen a ion is mos equen ly ecei ed. Resul s show ha in 73.74% o cases,
documen s a e ecei ed in pape o ma , as PDFs o images, while only 26.26% a e
ecei ed in s uc u ed digi al o ma s such as Excel, XML, o CSV. These esul s
p o ide insigh in o he po en ial o he adop ion o AI in accoun ing. AI ools in
accoun ing end o ope a e mos e icien ly when wo king wi h s uc u ed digi al da a
(e.g. Excel, XML, CSV). In con as , pape documen s, PDFs, and image iles equi e
addi ional con e sion p ocesses o become machine- eadable. Such p ocesses,
including Op ical Cha ac e Recogni ion (OCR), can be slow, less accu a e, and place
addi ional bu dens on he sys em. I may he e o e be concluded ha en e p ises need
o unde go bo h echnical and o ganisa ional ans o ma ion in o de o e ec i ely
in eg a e AI ools in o he accoun ing unc ion.
Beyond he o ma in which documen s a e ecei ed, ano he impo an dimension o
digi al eadiness conce ns he manne in which elec onic documen s a e a chi ed.
E icien a chi ing and da a accessibili y a e essen ial ounda ions o au oma ion and
he use o ad anced echnologies such as AI so wa e. To his end, he s udy also
examined how en e p ises manage he a chi ing o documen s ecei ed ia email.
Responses o his ques ion can se e as indica o s o he p esence—o absence—o
s uc u ed digi al documen managemen sys ems (such as DMS o cloud-based
solu ions). The dis ibu ion o esponses is p esen ed in Image 1.
Image 1.
Me hod o A chi ing Documen a ion Recei ed ia Email
Sou ce: Au ho ’s own analysis
The esul s indica e ha a signi ican p opo ion o en e p ises ely on basic documen
a chi ing me hods, such as physical s o age (37% plus 4% o hose awai ing
documen eceip by pos ) and s o age di ec ly wi hin email accoun s o on local d i es
(39%). These p ac ices hinde sys ema ic da a p ocessing and e ie al. Con e sely,
only 20% o esponden s u ilise ad anced documen managemen sys ems (DMS) o
cloud-based solu ions, which acili a e be e in eg a ion wi h AI ools. This si ua ion
highligh s he need o imp o e digi al in as uc u e o enable ull au oma ion o
accoun ing p ocesses.
To assess he au oma ion o business ansac ion eco ding, we in es iga ed whe he
he accoun ing so wa e used by i ms au oma ically sugges s accoun s o pos ing.
The esul s a e p esen ed in Image 2.
Image 2.
The capabili y o accoun ing so wa e o au oma ically sugges accoun s o pos ing
Sou ce: Au ho ’s own analysis
Au oma ic sugges ion o accoun s o pos ing o ms he ounda ion o as e and mo e
accu a e p ocessing o accoun ing documen a ion. This educes he possibili y o
human e o and p ocessing ime, while enabling employees o ocus on analy ical
and con ol asks.
Wi h he same pu pose o assessing he au oma ion o accoun ing eco ds, we also
in es iga ed he me hod o pos ing bank s a emen s. The esul s a e p esen ed in Image
3.
Image 3.
Me hods o pos ing bank s a emen s
Sou ce: Au ho ’s own analysis
The majo i y o esponden s (57%) manually ansc ibe da a om bank s a emen s,
while only 12% u ilise he op ion o impo iles ha a e au oma ically pos ed. An
38%
34%
28%
When you en e in oice da a, does you so wa e
au oma ically sugges accoun s o pos ing?
No
Yes
Unknown
57%
12%
6%
25%
How do you cu en ly pos bank s a emen s?
We manually ansc ibe he da a
We impo a ile ha is pos ed au oma ically
The so wa e au oma ically ecognises and pos s bank s a emen s
Unknown
e en smalle p opo ion, 6%, epo ha hei so wa e ecognises and pos s bank
s a emen s au oma ically, whe eas 25% o esponden s a e ei he unsu e o un amilia
wi h he pos ing me hod. These indings indica e a high deg ee o manual da a en y
in a c i ical segmen o he accoun ing p ocess, sugges ing ha businesses s ill lack
he adequa e echnical p e equisi es o he implemen a ion o AI ools in accoun ing
p ocedu es. Rega ding he digi al ma u i y o accoun ing—i.e., he ex en o which
digi al echnologies and AI a e in eg a ed in o accoun ing p ocesses—23.2% o
esponden s conside hemsel es o be in he ini ial phase (p ocesses a e
p edominan ly manual and lack s anda disa ion), 39.4% ega d hemsel es as being
in he second phase—digi ally awa e ( ecogni ion o he need o digi alisa ion and
basic ini ia i es ha e been launched), 27.3% iew hemsel es as being in he hi d
phase—digi ally ac i e (digi al ools in oduced, pa ial au oma ion), 8.1% iden i y
wi h he ou h phase—digi ally ad anced (in eg a ed sys ems and analy ics, high
e iciency), and only 2% conside hemsel es in he i h phase—inno a i e (use o
AI, p edic i e analy ics, wi h s a egic impac ). I can be concluded ha he cu en
s a e e lec s a low le el o digi alisa ion and au oma ion wi hin accoun ing p ocesses,
which cons i u es a signi ican ba ie o he e ec i e u ilisa ion o AI echnologies.
Howe e , he exp essed in en ion o some businesses o adop AI ools, alongside
exis ing basic o ms o digi alisa ion, indica es po en ial o u u e de elopmen . To
inc ease he adop ion o AI ools, a comp ehensi e echnical, o ganisa ional, and
educa ional ans o ma ion wi hin en e p ises is essen ial.
CONTRIBUTION OF AI TOOLS IN ACCOUNTING
The e alua ion o esponden s’ iews on he con ibu ion ha AI echnology can b ing
o accoun ing is p esen ed in Table 1. The esea ch esul s indica e ha esponden s
ecognise he posi i e po en ial o AI ools in imp o ing accoun ing p ocesses. The
highes a e age a ings we e awa ded o he a eas o au oma ion o pos ings (3.77)
and cos analysis (3.73), sugges ing ha hese a e pe cei ed as he mos sui able ields
o he applica ion o a i icial in elligence.
Table 1.
Responden s’ Views on he Con ibu ion o AI Tools in Accoun ing by Speci ic Segmen s
DESCRIPTION
AVERAGE
STDEV
Au oma ion o pos ings
3,77
1,25
P epa a ion o accoun ing epo s
3,65
1,21
Cos analysis
3,73
1,18
Managemen o ecei ables and payables
3,63
1,19
Cash low o ecas ing
3,45
1,21
E o de ec ion
3,62
1,17
Sou ce: Au ho ’s own analysis
Responden s’ iews highligh he ecognised alue o AI echnology in accoun ing,
pa icula ly in au oma ion and analy ical asks. Con e sely, he e emains scope o
u he awa eness- aising and educa ion ega ding mo e ad anced unc ionali ies such
as p edic ion and anomaly de ec ion.
The assessmen o esponden s’ opinions on wha would acili a e hei wo k o e he
nex 12 mon hs (in he sho - e m u u e) is p esen ed in Image 4.
Image 4.
Responden s’ pe cep ions o he needs o acili a ing accoun ing asks by segmen
Sou ce: Au ho ’s own analysis
The esea ch esul s indica e ha esponden s pe cei e he g ea es po en ial o
imp o ing hei wo k h ough he educ ion o manual da a en y in accoun ing. No
ewe han 67 esponden s iden i ied educed manual inpu as he key ac o ha
would mos acili a e hei daily asks o e he nex wel e mon hs. This is u he
complemen ed by he desi e o as e p ocessing o accoun ing documen a ion,
speci ically h ough he au oma ion o pos ing en ies. Responden s’ answe s clea ly
highligh he need o enhance e iciency by au oma ing undamen al accoun ing
p ocesses, pa icula ly hose in ol ing manual da a en y and p ocessing. When asked
whe he hey conside i impo an o ha e an AI ool in accoun ing ha independen ly
ecognises and p ocesses documen s, 47.5% eplied a i ma i ely, 48.5% conside ed
i pa ially impo an , and 4% s a ed i was no impo an . Con e sely, when
ques ioned whe he he ansi ion o au oma ic pos ing would educe hei con ol o e
da a, 26.3% esponded a i ma i ely, 31.3% said no, and 42.4% we e unce ain. These
indings sugges a cau ious a i ude among esponden s, unde sco ing he need o
u he educa ion and demons a ion o he secu i y and anspa ency o AI ools in
o de o o e come conce ns ela ed o po en ial loss o da a con ol.
In conclusion, he majo i y o esponden s ecognise he signi ican po en ial o
a i icial in elligence applica ion in accoun ing, pa icula ly in he a eas o pos ing
au oma ion and cos analysis. The esul s also e eal a s ong demand o educing
manual da a en y and accele a ing documen p ocessing, he eby con i ming a
eadiness o emb ace au oma ion o basic accoun ing asks. Howe e , he e emains a
deg ee o cau ion among esponden s ega ding ull au oma ion, la gely due o
conce ns o e loss o con ol o da a. This highligh s he necessi y o comp ehensi e
educa ion on he capabili ies, secu i y, and bene i s o AI ools in accoun ing
p ocesses, alongside awa eness o associa ed isks such as secu i y ulne abili ies and
echnological dependency.
67
55
41
28
010 20 30 40 50 60 70 80
Less manual da a en y
Fas e p ocessing o accoun ing
documen a ion
Au oma ion o epo p epa a ion
Au oma ion o clien communica ion (Cha bo )
Wha would make you wo k easie o e he nex 12 mon hs?
