Mas e Deg ee P og am in
Da a Science and Ad anced Analy ics
A i icial In elligence in he Banking Sec o : De elopmen o a
F amewo k o E ec i e Deploymen
Ana Ri a Figuei edo Ta a es
Mas e Thesis
p esen ed as pa ial equi emen o ob aining a Mas e ’s Deg ee in Da a Science and Ad anced Analy ics
NOVA In o ma ion Managemen School
Ins i u o Supe io de Es a ís ica e Ges ão de In o mação
Uni e sidade No a de Lisboa
MDSAA
NOVA In o ma ion Managemen School
Ins i u o Supe io de Es a ís ica e Ges ão de In o mação
Uni e sidade No a de Lisboa
A i icial In elligence in he Banking Sec o : De elopmen o a F amewo k o E ec i e
Deploymen
by
Ana Ri a Figuei edo Ta a es
Mas e Thesis p esen ed as pa ial equi emen o ob aining he Mas e ’s deg ee in Da a
Science and Ad anced Analy ics, wi h a specializa ion in Business Analy ics
Supe ised by
Ví o Dua e dos San os, PhD, NOVA IMS – In o ma ion Managemen School
July, 2024
i
STATEMENT OF INTEGRITY
I he eby decla e ha ing conduc ed his academic wo k wi h in eg i y. I con i m ha I ha e no
used plagia ism o any o m o undue use o in o ma ion o alsi ica ion o esul s along he
p ocess leading o i s elabo a ion. I u he decla e ha I ha e ully acknowledged he Rules
o Conduc and Code o Hono om he NOVA In o ma ion Managemen School.
Lisbon, 15.07.2024
ii
ACKNOWLEDGMENTS
I would like o begin by exp essing my p o ound g a i ude o my pa en s. I am e e nally
g a e ul o hei cons an suppo and belie in me and my ambi ions. Thei encou agemen
and lo e ha e been undamen al o my success and ha e se ed as he ounda ion o his
jou ney.
I am also g a e ul o acknowledge he in aluable suppo and encou agemen p o ided by my
boy iend, who has been a cons an sou ce o inspi a ion. I am immensely g a e ul o his
pa ience and unde s anding du ing his challenging jou ney.
My since es g a i ude ex ends o my iend Bea iz, whose assis ance, bo h echnical and
emo ional, was ins umen al in achie ing a posi i e ou come in my disse a ion.
I am ex emely g a e ul o my amily and iends, whose lo e and iendship ha e p o ided me
wi h he s eng h and de e mina ion o pe se e e. I am indeed o una e o be su ounded by
such admi able people, and I am e e nally hank ul o he joy and s abili y you b ing o my
li e.
I would also like o exp ess my acknowledgmen o my colleagues who pa icipa ed in his
disse a ion. Thank you o you ime, e o , and in aluable con ibu ions ha en iched he
quali y and dep h o his esea ch.
Finally, I would like o acknowledge he p i ilege I had o being able o wo k wi h P o esso
Ví o San os. His guidance and s ong wo k e hic we e ins umen al in di ec ing me owa d
he igh di ec ion, signi ican ly con ibu ing o my academic g ow h and success ul
de elopmen o his disse a ion.

iii
ABSTRACT
This disse a ion add esses he in eg a ion o A i icial In elligence (AI) in he banking sec o ,
wi h a pa icula ocus on he de elopmen o a s uc u ed amewo k o guide he sys ema ic
adop ion o AI echnologies, o e ing sys ema ic guidelines o assis he banking sec o in
iden i ying, e alua ing, and e ec i ely implemen ing AI echnologies. The esea ch employs
he Design Science Resea ch me hodology, s a ing wi h an ex ensi e e iew o exis ing
li e a u e on AI applica ions wi hin he banking indus y. Following his, he amewo k was
buil based on hese indings and e ined h ough he applica ion o he S a egic Alignmen
Model (SAM), ensu ing ha AI implemen a ions a e aligned wi h he s a egic objec i es o
banking ins i u ions. Subsequen ly, a su ey was conduc ed wi h wo indus y p o essionals,
dis inguished by hei le el o expe ise, o ga he insigh s and eedback, discussing he
limi a ions and sugges ions o u u e wo k, and highligh ing a eas o u he e inemen and
enhancemen o he amewo k. Despi e i s s eng hs, he ola ili y o he e olu ion o AI
echnology and he absence o mul iple use case demons a ions a e among he amewo k's
limi a ions. Consequen ly, u u e esea ch should ocus on ex ending he amewo k's
adap abili y o di e en banking en i onmen s since i could expand he amewo k's scope
o co e a wide ange o banking se ices, implemen i in a eal bank o u he e inemen ,
and explo e mo e use case scena ios o demons a e i s applicabili y and obus ness in di e se
con ex s.
KEYWORDS
A i icial In elligence; Banking Indus y; F amewo k De elopmen ; Inno a ion Managemen ;
Technology Adop ion
Sus ainable De elopmen Goals (SDG):
i
TABLE OF CONTENTS
1. In oduc ion .................................................................................................................. 1
1.1. Backg ound and p oblem iden i ica ion ................................................................ 1
1.2. Impo ance and ele ance .................................................................................... 2
1.3. Objec i es .............................................................................................................. 3
2. Me hodology ................................................................................................................ 4
2.1. Design Science Resea ch ....................................................................................... 4
2.2. DSR Implemen a ion ............................................................................................. 6
2.2.1. Rele ance Cycle .............................................................................................. 6
2.2.2. Rigo Cycle ...................................................................................................... 7
2.2.3. Design Cycle .................................................................................................... 7
3. Banking Indus y ........................................................................................................... 8
3.1. O e iew ................................................................................................................ 8
3.2. A eas ...................................................................................................................... 9
3.3. Oppo uni ies and Challenges ............................................................................. 11
3.3.1. Oppo uni ies ............................................................................................... 11
3.3.2. Challenges .................................................................................................... 13
4. Li e a u e Re iew ....................................................................................................... 15
4.1. A i icial In elligence ............................................................................................ 15
4.1.1. His o ical Backg ound................................................................................... 15
4.1.2. Concep s ....................................................................................................... 17
4.1.3. Technologies and Tools ................................................................................ 18
4.2. AI Technologies in he Banking Indus y ............................................................. 21
4.3. AI Challenges in Banking Indus y ....................................................................... 23
4.4. Assessmen o Economic Impac o Technologies .............................................. 25
4.4.1. S a egic Alignmen Model by Hende son and Venka aman ..................... 26
4.4.2. S a egic Alignmen Ma u i y Model by Lu man ........................................ 27
5. F amewo k P oposal ................................................................................................... 30
5.1. Assump ions ........................................................................................................ 30
5.2. P oposal ............................................................................................................... 31
5.2.1. Iden i ica ion o needs and objec i es ......................................................... 32
5.2.2. E alua ion and Selec ion o AI Technologies ............................................... 34
5.2.3. Design and In eg a ion ................................................................................. 35
5.2.4. Implemen a ion and Moni o ing .................................................................. 36
5.3. Use Case Demons a ion ..................................................................................... 36
5.3.1. Iden i ica ion o needs and objec i es ......................................................... 37
5.3.2. E alua ion and Selec ion o AI Technologies ............................................... 38
5.3.3. Design and In eg a ion + Implemen a ion and Moni o ing ......................... 39
5.4. E alua ion and Discussion ................................................................................... 40
6. Conclusions and u u e wo k ...................................................................................... 43
6.1. F amewo k Limi a ions ........................................................................................ 43
6.2. Fu u e wo k ......................................................................................................... 44
Bibliog aphical Re e ences .............................................................................................. 45
Appendix A ...................................................................................................................... 53
Appendix B ...................................................................................................................... 54
Appendix C ...................................................................................................................... 55
Annexes ........................................................................................................................... 60
i
LIST OF FIGURES
Figu e 2.1.1 - Design Science Resea ch ..................................................................................... 5
Figu e 2.2.1 - DSR Implemen a ion ........................................................................................... 6
Figu e 3.2.1 - Bank’s ope a ional s uc u e ............................................................................. 10
Figu e 4.1.1 - Illus a ion o AI de elopmen s o e ime ........................................................ 16
Figu e 4.1.2 - AI Sub ields ......................................................................................................... 20
Figu e 4.1.3 - Hie a chical S uc u e o AI Technologies ......................................................... 21
Figu e 4.3.1 - Numbe o AI inciden s and con o e sies, 2012-2021 .................................... 24
Figu e 5.2.1 - P oposed F amewo k ......................................................................................... 32
5
o use s, wi h he aim o enhancing he e ec i eness and e iciency o hese sys ems (He ne
e al., 2004).
Figu e 2.1.1 - Design Science Resea ch – Sou ce: (Pe e s e al., 2019) Adap ed.
Figu e 2.1.1 p o ides a isual and o e all in e p e a ion o he DSR P ocess, illus a ing he six
majo s eps in i s nominal sequence and highligh ing i s i e a i e na u e, allowing esea che s
o be lexible and s a in any s ep o p oduce imp o emen s (Pe e s e al., 2019).
The ini ial s age is pi o al, as i in ol es he iden i ica ion o he speci ic esea ch p oblem and
unde s anding i s impo ance, di ec ing he c ea ion o a iable a i ac and s imula ing he
esea che 's mo i a ion o pu sue he solu ion and emb ace he ou comes. Following he
ini ial iden i ica ion o he esea ch p oblem, he nex s ep in ol es se ing objec i es ha a e
ei he quali a i e o quan i a i e and equi e a ho ough unde s anding o he p oblem's
cu en s a e, he e ec i eness o exis ing solu ions, and an assessmen o wha is bo h
possible and easible. Subsequen ly, he design and de elopmen phase in ol es cons uc ing
he a i ac , based on de ined objec i es and ele an heo e ical amewo ks. Mo eo e , he
demons a ion phase es s he a i ac 's e icacy o sol e he p oblem in one o mo e speci ic
ins ances, le e aging a deep unde s anding o i s applica ion. E alua ion ocus on obse ing
and measu ing he a i ac 's pe o mance agains he se objec i es, which o en leads o
i e a i e e inemen s o enhance i s e icacy. Finally, he communica ion phase in ol es
a icula ing he signi icance o he p oblem and he u ili y o he a i ac , guided by i s
e ec i eness and me hodological igo (Pe e s e al., 2019).
In summa y, he diag am demons a es he lexibili y and adap abili y o he esea ch
me hodology, enabling esea che s o ini ia e om mul iple s a ing poin s. These include a
p oblem-cen e ed app oach, s a ing wi h he iden i ica ion o a c i ical issue; an objec i e-
cen e ed solu ion, beginning wi h a de ined goal; o a design and de elopmen -cen e ed
app oach, ocusing ini ially on a i ac c ea ion. This lexibili y suppo s di e se esea ch
s a egies and objec i es.

6
2.2. DSR IMPLEMENTATION
As highligh ed in he 2004 pape "Design Science in In o ma ion Sys ems Resea ch" by He ne
e al., design science has an impo an ole wi hin he IS discipline. I emphasizes he
impo ance o aligning IS esea ch wi h business needs and igo ously con ibu ing o he
knowledge base, he eby ensu ing he ele ance and igo o esea ch (He ne e al., 2004).
He ne ’s app oach o IS esea ch emphasizes main aining a balance be ween p ac ical
ele ance and heo e ical igo , aiming o bene i bo h business needs and he b oade
knowledge domain. This app oach is cha ac e ized by h ee in insic esea ch cycles: The
Design Cycle, which emphasizes he i e a i e de elopmen and assessmen o a i ac s; he
Rele ance Cycle, which ensu es ha esea ch is es ed in eal-wo ld scena ios; and he Rigo
Cycle, which bo h d aws om and con ibu es o exis ing heo e ical amewo ks and domain
knowledge, he eby enhancing he esea ch knowledge base (He ne , 2007).
Figu e 2.2.1 - DSR Implemen a ion - Sou ce: (He ne e al., 2004) Adap ed.
Figu e 2.2.1 isually demons a es ha by combining p ac ical ele ance wi h heo e ical
igo , he app oach e ec i ely add esses bo h business needs and knowledge ad ancemen
wi hin he IS discipline. This amewo k emphasizes a cyclical p ocess ha in eg a es he
en i onmen , he de elopmen and imp o emen o a i ac s, and he con ibu ion o bo h
business p ac ices and knowledge base.
2.2.1. Rele ance Cycle
This s age o DSR ini ia es by supplying he equi emen s o he esea ch as inpu s and
es ablishes he c i e ia o e alua ing he ou comes. Mo eo e , he esul s should be
ein oduced in o he en i onmen o examina ion and assessmen wi hin he applica ion
ield. Consequen ly, he cla i y o hese esul s de e mines whe he addi ional i e a ions a e
7
necessa y, pa icula ly i he e a e iden i ied de iciencies in unc ionali y o u ili y ha could
limi p ac ical applica ion. Fu he mo e, ano he i e a ion o he ele ance cycle ypically
begins wi h he eedback ecei ed om he en i onmen and he econ i ma ion o esea ch
equisi es de i ed om expe ience (He ne , 2007).
In his con ex , he en i onmen e e s o he banking sec o , including he oles, skills,
o ganiza ional amewo ks, p ocesses, and speci ic echnologies ela ed o AI implemen a ion.
This ocus ensu es he esea ch is aligned wi h p ac ical needs in banking.
2.2.2. Rigo Cycle
The Rigo Cycle le e ages a comp ehensi e and de ailed knowledge base, di ided be ween
he exis ing expe ise o he applica ion domain and he scien i ic heo ies ha unde pin
igo ous design science esea ch. This cycle ensu es ha inno a ion is in o med by es ablished
knowledge, demanding c ea i i y and he me iculous selec ion o igo ous s anda ds, he eby
suppo ing he de elopmen o au hen ic inno a ion while a oiding he supp ession o
c ea i e solu ions h ough inapp op ia e heo e ical cons ain s (He ne , 2007).
Rega ding his con ex , he knowledge base consis s o economic heo ies and AI's heo e ical
unde pinnings, me hodologies, and p io esea ch ha bo h in o ms and is en iched by
empi ical s udies wi hin he banking sec o . I emphasizes objec i i y o ensu e he e aci y
and dependabili y o indings, he eby en iching he scien i ic li e a u e wi h new insigh s. This
igo is i al o he c edibili y and ep oducibili y o he esea ch.
2.2.3. Design Cycle
The Design Cycle is cen al o DSR, ea u ing an i e a i e p ocess whe e a i ac s a e buil ,
e alua ed, and e ined. This cycle ope a es based on he equi emen s iden i ied in he
Rele ance Cycle and he heo ies p o ided by he Rigo Cycle, unc ioning wi h a deg ee o
independence and equi ing a balance be ween he cons uc ion and e alua ion o a i ac s,
ensu ing bo h p ocesses a e deeply g ounded in ele ance and igo . Mul iple i e a ions o
his cycle may be necessa y be o e he con ibu ions o he esea ch a e acknowledged wi hin
he Rele ance and Rigo Cycles (He ne , 2007).
8
3. Banking Indus y
To cons uc a de ailed amewo k ha cap u es and e alua es he ex ensi e impac o AI in
he banking sec o , i is undamen al o examine he indus y’s main componen s and p ima y
ope a ions.
This sec ion p o ides an examina ion o he undamen al aspec s o he banking indus y,
o e ing a comp ehensi e o e iew o he main ope a ional a eas and elucida ing he
signi ican challenges and oppo uni ies cu en ly acing banks. This analysis es ablishes he
ounda ion o he iden i ica ion o s a egic oppo uni ies o he deploymen o AI
echnologies ha may acili a e u he p og ess wi hin he sec o .
3.1. OVERVIEW
In he cu en economy, which places a s ong emphasis on cus ome se ices and cus ome
sa is ac ion, he banking indus y is ac i ely seeking ad ancemen s o imp o e i s business
p ocesses, enhance inancial p oduc s and se ices, and con ibu e o economic expansion
(Biswal, 2015). The e olu ion o his indus y due o compe i ion, complemen a y and co-
e olu ion (B oby, 2021).
Beginning wi h undamen al concep s, he unc ions o inancial ins i u ions include bo h
essen ial economic and inancial ope a ions and a g owing emphasis on enhancing se ice
quali y and imp o ing cus ome expe iences, which he e olu ion o digi al echnology has he
capaci y o change he na u e o banking (B oby, 2021). This holis ic app oach emphasizes he
essen ial ole o banks in d i ing economic de elopmen and hei commi men o mee ing
he changing needs o hei cus ome s, he eby inc easing cus ome sa is ac ion and p o i s
(Zoua i & Abdelhedi, 2021). The speci ic ange o ac i i ies pe o med depends on he
classi ica ion o he banking ins i u ion:
▪ Cen al Bank: his ins i u ion is pi o al in guiding and egula ing a na ion's banking
sys em. I manages deposi accoun s o all comme cial banks and o e s liquidi y
suppo as needed. Addi ionally, he Cen al Bank is esponsible o implemen ing
mone a y policy, egula ing he money supply, and ensu ing inancial s abili y
(Fede al Rese e, 2022).
▪ Comme cial Bank: his ype o bank p o ides a wide ange o inancial se ices,
along wi h o he essen ial banking unc ions. Al hough comme cial banks ca e o
bo h indi idual and co po a e clien s, hey ypically emphasize mee ing he
inancial equi emen s o business en i ies (Jiang, 2024).
▪ Re ail Bank: i may be de ined as a segmen o he inancial se ices indus y ha
ocuses on he in e media ion be ween indi idual consume s and inancial
p oduc s, ac ing as inancial in e media ies (Wu, 2023).
▪ In es men Bank: i is undamen al in suppo ing indi iduals, companies, and
go e nmen s in aising unds, unde w i ing, and acili a ing he issue o secu i ies.
9
Addi ionally, i ac s as an in e media y o in es o s in e es ed in he inancial
ma ke s, p o iding essen ial ansac ional suppo and s a egic in es men
guidance (Y. Liu, 2023).
The unc ions o a bank co e a ange o ac i i ies. I s co e unc ions include inancial se ices
such as ac ing as in e media ies be ween sa e s and bo owe s, g an ing loans wi h isk-
adjus ed in e es a es, and main aining deposi se ices. Beyond hese, banks a e also hea ily
in ol ed in p ocessing in o ma ion by e alua ing and managing he isks linked o hei asse
po olios. Mo eo e , mode n banking ope a ions a e signi ican ly cha ac e ized by
secu i iza ion ac i i ies (C. Wang, 2005).
The ope a ional me hods o banks ha e unde gone a p o ound ans o ma ion as a
consequence o he ad en o new echnologies (Roy e al., 2023). This echnological e olu ion
has ans o med he landscape o inancial se ices, o e ing clien s enhanced e iciency and
p omp ing ins i u ions o in es in cu ing-edge echnology and inno a ion o mee he
inc easing complexi y o demands (Mish a, 2015). The p ecision and accu acy o ansac ions
and inancial se ices ha e imp o ed, inc easing he accessibili y o in o ma ion and
con enience o cus ome s and, as a esul , banks a e o ced o inno a e and adap in o de
o main ain cus ome us and sa is ac ion, including he in oduc ion o new o ms o access
and inno a i e se ices ha a e inc easingly p e e ed o e adi ional in-pe son in e ac ions,
allowing cus ome s o comple e hei inancial ac i i ies e o lessly and con enien ly (B oby,
2021)
Consequen ly, he banking sec o aces ongoing challenges o imp o e i s business models,
aiming o con e obs acles in o oppo uni ies h ough he in eg a ion o echnological
inno a ions in o co e ope a ions (Ghandou , 2021). The success ul implemen a ion o AI
equi es a obus ounda ion in ou key a eas: o ganiza ional s uc u e, echnological
in as uc u e, p ocedu al amewo ks, and en i onmen al ac o s (Me hi, 2022). The e o e,
i manages mul i-disciplina y p ocedu es h ough collabo a ions ac oss he indus y.
Fu he mo e, he in eg a ion o echnology has spu ed he de elopmen o inno a i e
business models emphasizing digi aliza ion, enhanced secu i y, and pe sonalized se ices. This
e olu ion is d i en by he need o adap o dis up i e inno a ions ha a e eshaping he
adi ional app oaches o deli e ing alue o consume s (Temelko , 2020).
3.2. AREAS
The banking indus y is di ided in o wo p incipal a eas, he on o ice and he back o ice,
in o de o acili a e he smoo h ope a ional lows ha a e necessa y o he indus y o
unc ion e ec i ely allowing o specializa ion and e iciency gains. The on o ice is p ima ily
esponsible o cus ome in e ac ions, de e mining s a egies o a ac new cus ome s while
e aining exis ing ones, and p omo ing hei loyal y. The back o ice ocuses on in e nal
business p ocessing and logis ical asks, using his in o ma ion o e ine and enhance
p ocesses and ensu e e icien se ice deli e y (Zome dijk & V ies, 2007). The exchange o
10
in o ma ion be ween hese a eas c ea es a dynamic en i onmen ha p omo es mu ual
lea ning, leading o inno a i e solu ions and imp o ed ope a ional e iciency wi hin he bank
(Huang e al., 2021).
Fu he mo e, he wo di isions wi hin he banking indus y ope a e in a coo dina ed manne ,
p o iding a solid ounda ion o adop ing and in es ing in ad anced echnologies. This
in eg a ion acili a es he applica ion o AI ac oss di e en depa men s, ul ima ely aiming o
imp o e inancial e u ns and p omo e echnological ad ancemen wi hin he ins i u ion
(Huang e al., 2021).
Figu e 3.2.1 - Bank’s ope a ional s uc u e – Sou ce: (Huang e al., 2021) Adap ed.
Figu e 3.2.1 e ec i ely ep esen s he in e connec ion o he wo main blocks wi hin a bank,
highligh ing hei con inuous collabo a ion o achie e common goals and a homogeneous and
cohesi e sys em.
As s a ed in he sec ion 3.1, he a ie y o se ices and p ocesses wi hin he banking sec o is
in luenced by he classi ica ion o he ins i u ion. Howe e , he Ame ican P oduc i i y &
Quali y Cen e (APQC) p o ides comp ehensi e amewo ks ha enhance ope a ional
e iciency ac oss mul iple sec o s, by conduc ing ex ensi e benchma king s udies, esea ching
and publishing bes p ac ices, imp o ing p ocesses, and o e ing aining p og ams. I s P ocess
Classi ica ion F amewo k (PCF) o e s de ailed p ocesses ailo ed speci ically o nume ous
indus ies, including banking, and helps ins i u ions achie e g ea e ope a ional consis ency,
s eamline wo k lows, and enhance o e all pe o mance (Ame ican P oduc i i y & Quali y
Cen e (APQC), 2019).
In he banking sec o , his amewo k dis inguishes wo main ca ego ies: Ope a ional
P ocesses, and Managemen and Suppo P ocesses. Toge he , hese ca ego ies desc ibe
indus y p ocesses in de ail, co e ing a wide ange o asks om s a egic de elopmen and
p oduc managemen o isk managemen and IT suppo (APQC & IBM, 2020). This s uc u e
ensu es a holis ic app oach o managing and op imizing ope a ions.
Table 3.2.1 - P ocess Ca ego ies in he Banking Indus y – Sou ce: (APQC & IBM, 2020)
Adap ed.
Ope a ional P ocesses
Managemen and Suppo P ocesses
1. De elop Vision and S a egy
1. De elop and Manage Human Capi al

11
2. De elop and Manage P oduc s and
Se ices
2. Manage In o ma ion Technology
3. Ma ke and Sell P oduc s and Se ices
3. Manage Financial Resou ces
4. Sou ce and P ocu e Ma e ials and
Se ices
4. Acqui e, Cons uc , and Manage Asse s
5. Deli e Se ices
5. Manage En e p ise Risk, Compliance,
Remedia ion, and Resiliency
6. Manage Cus ome Se ices
6. Manage Ex e nal Rela ionships
7. De elop and Manage Business
Capabili ies
Table 3.2.1 p o ides an o e iew o he p ima y p ocess ca ego ies ou lined by he PCF,
speci ically o he banking indus y. Each ca ego y u he di ides in o mo e de ailed ac i i ies
ha ex ensi ely add ess bo h ope a ional and managemen unc ions essen ial o banking
ope a ions. PCF is s uc u ed in o i e hie a chical le els: Ca ego y, P ocess G oup, P ocess,
Ac i i y, and Task, each designa ed by a unique iden i ie (APQC & IBM, 2020). Fo simplici y
and cla i y, comp ehensi e desc ip ions up o he hi d le el a e included in Annex I.
3.3. OPPORTUNITIES AND CHALLENGES
A undamen al poin o acknowledge is ha banks mus mee he e ol ing demands o hei
cus ome s (Reliabili y, Tangibili y, Responsi eness, Assu ance, and Empa hy) o signi ican ly
in luence hei ope a ional s a egies (Zoua i & Abdelhedi, 2021). As cus ome expec a ions
e ol e and echnological ad ancemen s con inue o eme ge, banks a e equi ed o enhance
hei echnological capabili ies: his in e sec ion be ween changing consume needs and
echnological expansion demands banks o inno a e and e ol e, ensu ing hey emain
ele an , compe i i e, and p o i able in he ace o he de as a ing e ec s o digi al
ans o ma ion (B oby, 2021).
The e o e, i is pa amoun o banks o achie e an equilib ium be ween ad anced digi al
p ocesses and adi ional cus ome se ice o e ec i ely add ess he eme ging challenges o
secu i y, p i acy, and con iden iali y, among o he s. I is o i al impo ance o achie e his
equilib ium in o de o gua an ee cus ome sa is ac ion and loyal y in he digi al age (Biswal,
2015).
3.3.1. Oppo uni ies
The banking indus y p esen s a mul i ude o oppo uni ies o g ow h, inno a ion, and
compe i i eness. These oppo uni ies a e d i en by a numbe o s a egic ini ia i es, including
he adop ion o In e ne echnology, apid echnological ad ancemen s, de egula ion,
globaliza ion, and changes in compe i i e and egula o y en i onmen s (Subbu aj, 2023).
Addi ionally, digi al ans o ma ion ac s as a undamen al ca alys , c ea ing oppo uni ies o
inno a ion ia digi al channels, he eby enhancing cus ome expe iences, accessing
12
unexplo ed sec o s, and s eamlining ope a ional p ocedu es. The e o e, his unde sco es he
necessi y o banks o adop digi al ans o ma ion o main ain hei ele ance and
compe i i eness (Ghandou , 2021).
Fin ech, which s ands o inancial echnology, e e s o he applica ion o inno a i e
echnological solu ions o imp o e and au oma e inancial se ices. The banking sec o is
unde going signi ican digi al ans o ma ion, which is hea ily in luenced by in ech
inno a ions, which a e eshaping adi ional business models (Temelko , 2020). The e o e,
banks ha es ablish pa ne ships wi h in ech companies may adop hese ad anced
echnologies and inno a i e business models, which e olu ionize adi ional banking
p ac ices (Josyula, 2021). In summa y, his echnology enables inancial ins i u ions o mo e
e icien ly p o ide inancial se ices, wi h g ea e lexibili y, he eby enabling banks o employ
ad anced echnologies and inno a i e business models, ex ending hei ange o se ices and
inc easing ope a ional e iciency (Temelko , 2020).
Fu he mo e, he e is a subs an ial oppo uni y o he s a egic u iliza ion o knowledge
managemen (KM) as a business p ac ice, gi en ha i s e ec i e adop ion may esul in a mo e
a o able e u n on in es men . While ypically linked wi h IT depa men s, he bene i s o
knowledge managemen ex end o se e al a eas o banking ope a ions due o i s abili y o
app oach compe i i e ad an age (Jayasunda a, 2008). In summa y, he e is empi ical
e idence o he powe o KM in enhancing o ganiza ional compe i i eness and e iciency,
leading o enhanced p ocess and employee pe o mance, posi i ely a ec ing ma ke
pe o mance and, ul ima ely, he o ganiza ion's o e all success (Cebi e al., 2010).
Mo eo e , acknowledging he impo ance o sus ainable banking o e s signi ican
oppo uni ies o g ow h and inno a ion in he inancial sec o , enabling banks o access new
ma ke segmen s ocused on g een echnologies and en i onmen ally conscious cus ome s
while enhancing hei epu a ion h ough a commi men o sus ainabili y. In addi ion,
sus ainable banking p ac ices may imp o e isk managemen by an icipa ing and mi iga ing
he impac s o clima e change (De Haas, 2023). I is inc easingly seen as a s a egic ma ke
ad an age, including in eg a ing en i onmen al conside a ions in o ope a ions, de eloping
g een inancial p oduc s, and o mula ing s a egies o add ess clima e change, while pu suing
hei p o i -making ac i i ies. The e o e, by in eg a ing sus ainabili y guidelines in o he
s a egic, inancial, and ope a ional decision-making p ocesses, banks may signi ican ly
suppo en i onmen ally and socially esponsible p ojec s, enabling hem o play a pi o al
posi ion in he p omo ion o sus ainable de elopmen (Ca è, 2018).
In u n, globaliza ion p esen s he banking indus y wi h se e al oppo uni ies, including he
expansion in o o eign ma ke s, di e si ica ion o se ices, he in eg a ion o echnology,
ma ke -d i en inno a ions, and he o ma ion o s a egic alliances (R.K. & Rimpi, 2006).
Globaliza ion, coupled wi h he e olu ion o cus ome p e e ences, he pu sui o economies
o scale, di e si ica ion o co e business ac i i ies, egula o y changes, and echnological
ad ancemen s, has led o he ad en o new business models (Temelko , 2020).
13
In conclusion, he banking landscape is cha ac e ized by di e si y and cons an e olu ion. By
adop ing a s a egic app oach, banks may e ec i ely mee consume expec a ions, demands,
and equi emen s, ensu ing hey main ain hei unique iden i y while adap ing o changes
(Subbu aj, 2023).
3.3.2. Challenges
In he apidly e ol ing banking indus y, banks ace a mul i ude o complex challenges as hey
a emp o adjus o con inual ans o ma ions. These challenges include cus ome
sa is ac ion and e en ion, compe i ion, he deploymen o adequa e echnology, egula o y
obs acles, and o he s (Biswal, 2015).
The banking indus y aces nume ous egula o y challenges ha signi ican ly a ec i s
ope a ions and in eg i y. The lack o obus egula o y amewo ks agg a a es hese issues,
allowing audulen ac i i ies o pe sis . In o de o e ec i ely add ess hese egula o y
challenges, i is necessa y o he indus y o implemen comp ehensi e e o ms, imp o e i s
echnological in as uc u e, and enhance collabo a ion among inancial ins i u ions (Roy e
al., 2023).
Nowadays, banks a e cen e ing hei ope a ions a ound cus ome needs, wi h hei
expec a ions and demands guiding he indus y's di ec ion. The di e se ange o cus ome
pe spec i es and p o iles demands highly pe sonalized se ices, undamen al o main aining
cus ome loyal y. Howe e , adap ing o hese needs is becoming inc easingly challenging due
o he apid e olu ion o cus ome p e e ences (Ind iasa i & Gaol, 2019). The p ocessing o
la ge olumes o da a ep esen s a signi ican and complex challenge o he banking sec o
and inapp op ia e use o such da a may lead o egula o y challenges o banks, including da a
p i acy, algo i hmic anspa ency, and he e ol ing egula o y landscape (Ahmadi, 2024).
Fu he mo e, accu a ely measu ing and assessing compe i ion poses a signi ican challenge
o inancial ins i u ions, which mus con inuously adap o keep up wi h he apid e olu ion
o he sec o (B oby, 2021).
The eme gence o al e na i e inancial se ices p o ide s, such as digi al banks and neobanks,
has in ensi ied compe i ion in he banking indus y. These new en i ies challenge adi ional
banks o c ea e mo e use - iendly and accessible inancial p oduc s. This inc eased
compe i ion, especially om echnology-o ien ed ins i u ions, emphasizes he c i ical need
o adi ional banks o inno a e and emb ace digi al ans o ma ion in o de o emain
ele an and compe i i e (Temelko , 2020). To main ain compe i i eness, banks mus
inno a e hei p oduc o e ings and enhance cus ome se ice expe iences (Biswal, 2015).
The banking indus y aces signi ican social and e hical challenges, pa icula ly in he domain
o cus ome ca e. These challenges a e mul i ace ed and ha e signi ican implica ions o
us , anspa ency, da a secu i y and p i acy, egula o y compliance, and e osion o
epu a ion. In o de o add ess hese challenges, banks mus es ablish a cul u e o e hical
14
beha io , ensu e anspa en communica ion, p o ec cus ome da a, and p ac ice ai lending
(Sama a hunga & Ka una hilaka, 2023).
As digi al banking pla o ms expand and elec onic ansac ions become mo e p e alen ,
cybe secu i y has eme ged as a majo issue o he inancial indus y (Zoua i & Abdelhedi,
2021). The analysis and managemen o la ge olumes o sensi i e inancial da a in oduces
signi ican isks ha may comp omise cus ome p i acy, damage he epu a ion o inancial
ins i u ions, and h ea en hei o e all s abili y. Fu he mo e, he p o ec ion o da a is o
pa amoun impo ance in o de o p e en subs an ial inancial losses and se e ely diminish
consume us (Abioye e al., 2021). E ec i e isk managemen is essen ial o banks, as hey
ace a ious h ea s o hei s abili y, p o i abili y, and epu a ion, including ope a ional,
compliance, and echnological isks. Mo eo e , isk managemen aims o p e en losses and
o c ea e a s able and secu e en i onmen ha ein o ces us , mee s egula o y s anda ds,
and suppo s in o med decision-making. Addi ionally, banks mus e icien ly iden i y, assess,
and moni o hese isks o mi iga e po en ial ad e se e ec s and main ain a compe i i e edge
(Al-Tamimi, 2007). This equi es implemen ing comp ehensi e s a egies ha combine
ad anced analy ics, igo ous in e nal con ols, and a s ong cul u e o isk awa eness
h oughou he o ganiza ion (Al-Tamimi, 2007).
Ul ima ely, imp o ing ope a ional e iciency in banks equi es op imizing p ocesses, managing
esou ces e ec i ely, and u ilizing echnology o inc ease p oduc i i y, lowe cos s, and
imp o e se ice quali y. The inhe en complexi ies and occasional edundancies in banking
p ocedu es pose signi ican challenges and add essing hese e ec i ely demands a solid
s a egy, in es men in ad anced echnologies, and a deep commi men o o ganiza ional
enhancemen . Op imizing ope a ional e iciency is i al o banks o emain compe i i e,
adjus o ma ke luc ua ions, and sa is y cus ome demands (Dumasiya, 2023).
In conclusion, by p oac i ely add essing hese mul i ace ed challenges, banks may ensu e
hei con inued e iciency in he cons an ly changing global inancial en i onmen (Dumasiya,
2023).
21
Figu e 4.1.3 - Hie a chical S uc u e o AI Technologies - Sou ce: (Zhuhada & Ly as, 2023)
Adap ed.
As illus a ed in Figu e 4.1.3, his s uc u ed p og ession om AI o machine lea ning, hen o
deep lea ning, and inally o gene a i e AI illus a es an inc easing le el o specializa ion and
capabili y in which each laye builds on he echnologies and me hodologies o he p e ious
laye s and e ines hem, leading o mo e ad anced and specialized applica ions and
inno a ions (Zhuhada & Ly as, 2023). Such hie a chical s uc u ing is undamen al o he
de elopmen o inc easingly sophis ica ed AI applica ions capable o dealing wi h complex
asks.
These ools and echnologies play a pi o al ole in he de elopmen o AI sys ems, d i ing he
apid g ow h o he ield and signi ican ly inc easing i s impac ac oss mul iple sec o s. In o de
o achie e g ea e ou pu om hem, banks mus le e age hei capabili y o ha ness such
ad anced echnologies ully, maximizing AI's po en ial as a d i e o economic g ow h and
indus y ans o ma ion (Di ya, 2024).
4.2. AI TECHNOLOGIES IN THE BANKING INDUSTRY
As p e iously s a ed, he po en ial o AI o d i e economic g ow h is conside able, and i is
an icipa ed ha i will enhance decision-making p ocesses and acili a e con inuous
moni o ing o economic banking ac i i y based on a mo e comp ehensi e ange o
in o ma ion (Go indha aj, 2024).
Consequen ly, he p o ound impac o AI is signi ican ly de e mined by he ex en o which
companies adop i and how well s akeholde s a e p epa ed ac oss in ellec ual, echnological,
poli ical, e hical, and social dimensions. This comp ehensi e p epa a ion allows he banking
indus y o ma ch echnological de elopmen s by di e si ying and enhancing i s p oduc
o e ings h ough inc eased echnology in eg a ion and au oma ion (Go indha aj, 2024).
The in eg a ion o ML in banking has he po en ial o signi ican ly enhance au oma ion,
analysis, and decision-making p ocesses, p o iding a new way o mee cus ome s’ demands.
Fu he mo e, his echnology plays a pi o al ole in he main enance and enhancemen o
cus ome expe iences and us h ough a p o ound comp ehension o hei p e e ences and

22
beha io s (Donepudi, 2017). Consequen ly, banks may o e pe sonalized se ices, imp o e
hei in es men s a egies, and gain new insigh s in o he compe i i e landscape (Sahu e al.,
2023). Fu he mo e, machine lea ning enhances ope a ional e iciency by au oma ing ou ine
asks and p ocesses, he eby educing he eliance on manual labo . Addi ionally, i acili a es
he managemen o isk by accu a ely e alua ing bo owe isks and s eamlining loan
decision p ocesses. In summa y, ML ools acili a e he de ailed and p ecise analysis o
ex ensi e business and cus ome - ela ed da ase s, combining a ious da a elemen s o
ex ac aluable insigh s (Ghandou , 2021).
Subsequen ly, he ield o deep lea ning has been demons a ed o ha e ex ensi e
applica ions ac oss a ange o domains, including compu e ision, na u al language
p ocessing, speech ecogni ion, and ecommenda ion sys ems (Olaoye & Po e , 2024). In he
banking sec o , his echnique is employed o a a ie y o inno a i e applica ions, including
he analysis o ansac ion pa e ns o iden i y aud and he use o p edic i e analy ics o
accu a ely o ecas cus ome beha io and ma ke ends (Ghandou , 2021). Consequen ly,
his indus y is one o he as es expanding his sophis ica ed o m o ML (Go indha aj, 2024),
in o de o achie e subs an ial ad ancemen s in a ela i ely sho pe iod.
Fu he mo e, he u iliza ion o NLP and cha bo s is becoming inc easingly p e alen in he
inancial se ices indus y due o he ac ha i assis s banks in o e coming adi ional
challenges by enhancing cus ome se ice and au oma ing ou ine asks: his echnology
s eamlines in e ac ions, enabling AI o play a pi o al ole in educing cus ome se ice ime
and inc ease cus ome sa is ac ion (Ridha & Maha ani, 2022). Ne e heless, as banks adop
his echnology, i is pa icula ly impo an o ensu e i enhances hei se ices in o de o
e ain cus ome s, since he educed human con ac , high expec a ions om cus ome s, and
he high a e o pe cei ed isks may no gua an ee highe se ice quali y and po en ially e ode
cus ome loyal y (Ga la, 2018). Fu he mo e, his echnology au oma es he cus ome se ice
p ocess, eco ds a subs an ial amoun o cus ome da a, p o ides pe sonalized esponses and
ecommenda ions, ensu es he deli e y o accu a e and c edible communica ions, and
enhances cus ome sa is ac ion (Sa i & Adinda, 2023). Consequen ly, i signi ican ly enhances
he o e all cus ome expe ience.
Addi ionally, he implemen a ion o AI-based sys ems enables banks o au oma e asks,
educe e o s, and pe sonalize cus ome expe iences, allowing banks o s eamline hei
ope a ions and be e se e hei ech-sa y cus ome s. Addi ionally, AI's abili y o analyze
as amoun s o da a empowe s banks o make da a-d i en decisions in a eas such as isk
managemen and loan app o als (Donepudi, 2017).
Fu he mo e, a s udy by Pa ick Ul ich and Vanessa F ank emphasizes how his o e o
oppo uni ies and ools o p ocess au oma ion and e icien da a u iliza ion leads o an
accele a ion o p ocesses, po en ial sa ings, and ul ima ely leads o enhanced decision-
making. Addi ionally, he de elopmen o new business models and imp o emen s in isk
23
managemen a e signi ican ou comes o his echnological ad ancemen (Ul ich & F ank,
2021).
These inno a i e ools ha e he po en ial o in eg a e wi h he co e componen s o he
banking sec o , he eby impac ing he wo banking blocks p e iously men ioned in dis inc
ways. Fi s ly, di ec cus ome s a e inc easingly managed by AI, no ably h ough cha bo s o
i ual agen s, leading o a wide ange o bene i s, including enhanced ad ice, ailo ed o e s,
and ime sa ings. Finally, he back-o ice le e ages AI o iden i y anomalies and excep ions,
while a emp ing o mi iga e isks, such as o e - eliance on AI and he cos s associa ed wi h
implemen ing ini ia i es. In his g oup di ec cus ome bene i s migh no be immedia ely
appa en , ye i plays a c i ical ole in h ea de ec ion and isk mi iga ion (Lakhangaonka &
Kama h, 2021).
In summa y, he inancial sec o is inc easingly adop ing AI echnologies and implemen ing
dynamic and inno a i e models ha a e employed o accu a ely enhance isk managemen ,
enhance cus ome expe ience, and sus ain compe i i eness (S oboda, 2023).
4.3. AI CHALLENGES IN BANKING INDUSTRY
The apid ad ancemen o ad anced AI and da a science echnologies in ecen yea s has
c ea ed a new e a o echnological oppo uni ies in he inancial sec o , wi h applica ions
ex ending o a wide ange o inancial domains. Howe e , i has also in oduced nume ous
challenges and unce ain ies (Ness & Muhammad, 2024): he in eg a ion o AI in o banking
ope a ions is accompanied by a se ies o challenges, including conce ns ega ding da a
p i acy, e hical conside a ions, he need o comply wi h egula o y equi emen s, and he
po en ial o job displacemen . Fu he mo e, as AI echnologies con inue o ad ance and
e ol e, he conside a ions su ounding hei de elopmen become inc easingly complex,
po en ially h ea ening con en ional business s a egies and p o i abili y (S oboda, 2023).
The 2023 AI Index Repo has iden i ied ha he numbe o AI inciden s and con o e sies
wi hin di e en sec o s has inc eased 26 imes since 2012, as can be con i med by analyzing
he Figu e 4.3.1. This expansion is indica i e o g ea e use o AI echnologies and g owing
ecogni ion o he po en ial o misuse (S an o d Ins i u e, 2023).
24
Figu e 4.3.1 - Numbe o AI inciden s and con o e sies, 2012-2021 – Sou ce: (S an o d
Ins i u e, 2023) Adap ed.
The challenges in he banking indus y can be agg ega ed in o echnical, egula o y, e hical,
and socie al (Akh e e al., 2024). The ini ial challenge includes obs acles om he
managemen o da a, which is in luenced by bo h he quan i y and quali y o he da a and
may esul in a po en ial lack o quali y, as he p edic ion powe o an algo i hm may be
comp omised (Mhlanga, 2020). Fu he mo e, he lack o adequa e quali y da ase s may also
signi ican ly a ec he quali y and us wo hiness o he AI models being ained (Ghandou ,
2021). Addi ionally, inancial cons ain s, de iciencies in ins i u ional in o ma ion echnology
in as uc u es, and a lack o echnical expe ise p esen addi ional challenges o he e ec i e
implemen a ion o echnologies (Ul ich & F ank, 2021). An excessi e eliance on AI o
au oma e decision-making and p oblem-sol ing p ocesses migh also supp ess wo ke s'
c ea i i y and adap abili y, he eby limi ing he dynamic capabili ies o he wo k o ce
(Ghandou , 2021). The issue o da a secu i y and p i acy, as well as anspa ency in he
algo i hms, emains a signi ican egula o y conce n o he sec o (Ahmadi, 2024).
Consequen ly, he absence o de ini i e egula o y guidelines complica es he implemen a ion
o AI solu ions in he banking indus y, equi ing con inuous and mul i ace ed app oaches (AL-
Dosa i e al., 2022). Fu he mo e, he banking sec o is subjec o isk-o ien ed egula ions
ha may p o e inadequa e in ecognizing he dis inc i e ci cums ances o ulne able g oups,
gi en ha he use o new kinds o da a in oduces new p i acy and da a secu i y issues
(Mhlanga, 2020). Mo eo e , he de elopmen o in elligen sys ems p esen s a signi ican
challenge in achie ing high anspa ency and in e p e abili y, as i may lead o a lack o us
and con idence in hei ou pu s, which may ha e de imen al e ec s on he c edibili y o he
sys em (Wo ham e al., 2016). Complemen a y, he eme gence o AI-d i en banking b ings
signi ican socie al challenges. Fi s ly, he a ying le els o echnological accessibili y ac oss
di e en social g oups con ibu e o and shape he unequal dis ibu ion o hese ad anced
se ices: he challenge o digi al inancial inclusion is a signi ican issue o banks (X. Wang &
He, 2020). Simila ly, he e a e conce ns abou he po en ial impac o hese echnologies on
25
exis ing employmen pa e ns. In pa icula , he e is a ea ha hey could esul in he
obsolescence o ce ain skills, which in u n could lead o he displacemen and loss o jobs
(Ghandou , 2021).
Fu he mo e, as p e iously s a ed, while i is e iden ha AI has he po en ial o enhance
business p oduc i i y, i s ul ima e impac on g ow h emains dependen on he e ec i eness
o i s implemen a ion (Oluwaseyi & Po e , 2024). This complexi y se es o highligh he
ele ance o William J. Baumol and William G. Bowen's concep o "Baumol's cos disease",
which was o mula ed in he 1960s. This concep s a es ha "g ow h can be limi ed no by
wha we a e good a , bu a he by wha is essen ial and ye di icul o imp o e". E en in he
e en o au oma ion eplacing a signi ican numbe o oles, g ow h may s ill be cons ained
by a eas ha emain essen ial bu a e di icul o imp o e (Jones e al., 2017).
In o de o ully le e age he inno a i e po en ial o his echnology and o success ully
implemen AI, banks mus adop a comp ehensi e app oach ha add esses he challenges
e e enced abo e ha limi he deploymen o AI echnologies h oughou he ins i u ion
(B igh wood, 2024):
▪ Clea ly de ine he business objec i es;
▪ De elop a obus da a s a egy;
▪ Collabo a e wi h echnology pa ne s;
▪ S a wi h small-scale P oo -o -Concep (PoC) p ojec s o alida e AI solu ions and
assess hei e ec i eness;
▪ E hical conside a ions;
▪ Implemen a change managemen s a egy o suppo he cul u al shi ;
▪ Ensu e compliance wi h egula o y amewo ks and indus y s anda ds;
▪ Con inuous moni o ing and e alua ion;
▪ Plan o u u e g ow h and expansion o AI capabili ies.
In summa y, he sys ema ic li e a u e has demons a ed ha al hough AI o e s signi ican
oppo uni ies o he banking sec o , he esul ing challenges pe sis en ly in luence a ious
aspec s o he indus y. Consequen ly, ensu ing he p ope and success ul in eg a ion o AI
in o he banking sec o depends on ackling hese challenges (S oboda, 2023).
4.4. ASSESSMENT OF ECONOMIC IMPACT OF TECHNOLOGIES
Assessing he economic impac o such echnologies equi es a comp ehensi e app oach and
u he analysis, gi en he limi ed numbe o comp ehensi e s udies ha es ablish causal
ela ionships be ween AI sys ems and hei e ec s, highligh ing he u gen need o mo e in-
dep h esea ch in his ield.
The main componen s o a comp ehensi e echnology assessmen , acco ding o Joseph F.
Coa es and his esea ch, a e based on he ollowing p inciples (Coa es, 1974):
1. Examine p oblem s a emen s;
2. Speci y sys ems al e na i es;
26
3. Iden i y possible impac s;
4. E alua e impac s;
5. Iden i y he decision appa a us;
6. Iden i y ac ion op ions o he decision appa a us;
7. Iden i y pa ies o in e es ;
8. Iden i y mac o sys em al e na i es (o he ou es o goal);
9. Iden i y exogenous a iables o e en s possibly in luencing;
10. Conclusions (and ecommenda ions).
This app oach ensu es a holis ic unde s anding o he economic impac o echnologies,
helping o de elop s a egies ha maximize bene i s while minimizing nega i e impac s
(Coa es, 1974).
The oles o IT a chi ec s and hei impo ance o IT-business alignmen a e s ill unde a ed in
bo h heo y and p ac ice. Howe e , i is known ha he p incipal objec i es o IT-Business
alignmen a e o enhance o ganiza ional pe o mance, including educing cos s, inc easing
e enues, and imp o ing e u ns on in es men , o imp o e quali y and o os e posi i e
eac ions o eme ging oppo uni ies, and o gain a compe i i e ad an age h ough IT
(Gellweile , 2022). This usion enables he alignmen among a ious business unc ions,
enhancing bo h e ec i eness and e iciency and e ol ing in o a dynamic ela ionship whe e
IT and o he business sec o s collabo a e o adap hei s a egies (Lu man, 2001).
Fu he mo e, i is achie ed h ough he equi emen o an ongoing e o o s a egic planning,
ealignmen o objec i es, and implemen a ion o bes p ac ices o suppo and de ine
business s a egies (Chen, 2010).
The S a egic Alignmen Model (SAM) by Hende son and Venka aman, and he S a egic
Alignmen Ma u i y Model (SAMM) by Lu man a e undamen al o aligning IT amewo ks
and applica ions wi h business s a egies in he digi al banking landscape. While SAM p o ides
a model o esea ch and p ac ice o s a egic managemen o in o ma ion echnology
(Hende son & Venka aman, 1994), Lu man's model o e s a p ac ical me hod o e alua e
and enhance his alignmen 's ma u i y o e ime (Lu man, 2001). Bo h models will be
discussed in de ail in he subsequen sec ions.
4.4.1. S a egic Alignmen Model by Hende son and Venka aman
By ecognizing he limi a ions o exis ing amewo ks in p o iding undamen al knowledge and
guidance, Hende son and Venka aman in oduced hei S a egic Alignmen Model in 1994,
ma king he beginning o a new e a in esea ch on he subjec . The p oposed model is based
on ou main domains: business s a egy, o ganiza ional in as uc u e and p ocesses, IT
s a egy, and IT in as uc u e and p ocesses (Hende son & Venka aman, 1994).
The concep ual amewo k sugges s ha by in eg a ing bo h ex e nal and in e nal business
componen s, companies may e ec i ely synch onize hei IT and business objec i es, he eby
achie ing compe i i e ad an age and ope a ional excellence. Addi ionally, SAM de ines hese

27
componen s in o wo key bi a ia e i ela ionships: s a egic i which e e s o he alignmen
be ween ex e nal and in e nal componen s o he o ganiza ion (s a egic le el), and unc ional
in eg a ion e e ing o alignmen be ween business and IT domains wi hin he o ganiza ion
(ope a ion le el) (Hende son & Venka aman, 1994).
The model p oposes ha he in eg a ion o c oss-domain pe spec i es, which includes hese
dimensions, is mo e e ec i e han any indi idual bi a ia e i ela ionship in enhancing he
s a egic IT managemen e ec i eness (Hende son & Venka aman, 1994).
4.4.2. S a egic Alignmen Ma u i y Model by Lu man
Je y Lu man emphasized ha unde s anding he alignmen ma u i y o an o ganiza ion and
aking he necessa y ac ion o imp o e he IT-business ha mony is pi o al (Lu man, 2001).
The p ima y objec i e o his amewo k is o p o ide o ganiza ions wi h a ool o assess he
ma u i y o hei s a egic choices and alignmen ac i i ies, he eby iden i ying oppo uni ies
o imp o e s a egic cong uence. In addi ion, he model p o ides a mechanism o assessing
he le el o ma u i y o s a egic decisions and alignmen e o s, iden i ying a eas o
po en ial imp o emen (Chen, 2010).
In acco dance wi h Lu man's concep ualiza ion, s a egic alignmen includes wel e elemen s
dis ibu ed ac oss ou domains: Business S a egy, O ganiza ion In as uc u e and P ocesses,
IT S a egy, and IT In as uc u e and P ocesses (Chen, 2010). Each domain is comp ised o
h ee key componen s (Lu man, 2001):
▪ Business S a egy:
▪ Business Scope: Includes ma ke s, p oduc s, se ices, cus ome g oups, and
geog aphic loca ions whe e he company compe es, as well as compe i o s
and po en ial compe i o s ha in luence he business en i onmen ;
▪ Dis inc i e Compe encies: Includes c i ical success ac o s and co e
compe encies ha p o ide he i m wi h a po en ial compe i i e edge;
▪ Business Go e nance: Focuses on he ela ionships be ween managemen ,
s ockholde s, and he boa d o di ec o s.
▪ O ganiza ion In as uc u e and P ocesses:
▪ Adminis a i e S uc u e: he manne in which he i m s uc u es i s
business ope a ions;
▪ P ocesses: he manne in which he i m’s business ac i i ies a e conduc ed
and p oceed;
▪ Skills.
▪ IT S a egy:
▪ Technology Scope: ou lines he key in o ma ion applica ions and
echnologies;
▪ Sys emic Compe encies: hose capabili ies ha dis inguish IT se ices om
one ano he ;
28
▪ IT Go e nance: conce ned wi h he de ini ion o oles and esponsibili ies
o IT esou ces.
▪ IT In as uc u e and P ocesses:
▪ A chi ec u e: he ounda ion o c ea ing a cohesi e and ope a ional IT
sys em;
▪ P ocesses: ac i i ies o ope a ions pe o med wi h he objec i e o
de eloping and main aining applica ions and managing he associa ed IT
in as uc u e;
▪ Skills.
Fu he mo e, he s a egic alignmen is based on six IT-Business ma u i y ca ego ies (Lu man,
2001).
1. Communica ion Ma u i y: his ca ego y e alua es he issues o le e aging
in o ma ion o mu ual unde s anding and knowledge sha ing be ween business
and IT;
2. Compe ency/Value Measu emen Ma u i y: i e e s o he capaci y o IT
o ganiza ions o demons a e hei alue in a manne ha is comp ehensible and
accep able o he business;
3. Go e nance Ma u i y: i e e s o he s uc u ed p ocess whe e business and IT
leade s o mally discuss and p io i ize IT esou ces, ensu ing alignmen . In o de o
manage and alloca e esou ces e ec i ely, i is necessa y o ha e clea ly de ined
decision-making au ho i y;
4. Pa ne ship Ma u i y: i e e s o he collabo a i e ela ionship be ween business
and IT o ganiza ions and emphasizes he impo ance o IT ha ing an equal ole in
de ining business s a egies, p omo ing mu ual us , ha ing business sponso s and
champions o IT ini ia i es, and sha ing isks and ewa ds;
5. Scope & A chi ec u e Ma u i y: i assesses he capaci y o IT o acili a e he
implemen a ion o a lexible and anspa en in as uc u e ha se es all business
pa ne s and cus ome s. This includes he capabili y o ex end beyond he on
and back o ices, he e ec i e applica ion o eme ging echnologies, he acili a ion
o d i ing o business p ocesses and s a egies, and he p o ision o cus omizable
solu ions;
6. Skills Ma u i y: his e alua es he human esou ce aspec , pa icula ly he le el o
IT and business skills ac oss he o ganiza ion, owa d change and inno a ion. I
includes an e alua ion o he o ganiza ion's eadiness o change, he pe sonal
esponsibili y indi iduals eel o business inno a ion, he abili y o lea n om
expe iences apidly, and he capaci y o le e age inno a i e ideas and
en ep eneu ship.
Complemen a ily, he model in ol es i e le els o s a egic alignmen ma u i y ha , in u n,
desc ibes he ca ego ies men ioned p e iously (Lu man, 2001):
29
1. Ini ial/Ad Hoc P ocess: O ganisa ions ha mee his le el a e ca ego ized by he
minimal alignmen be ween IT and business s a egies and by he highly
imp obable likelihood o achie ing an aligned IT business s a egy. O ganiza ions
a his le el lack o malized p ocesses o aligning IT ini ia i es wi h business
objec i es.
2. Commi ed P ocess: a his le el, o ganiza ions begin o ecognize he impo ance
o alignmen and commi esou ces owa ds i . The e o o es ablish mo e
s uc u ed p ocesses be ween IT and business uni s may s ill be inconsis en and
challenging o achie e.
3. Es ablished Focused P ocess: a his hi d le el, he business has al eady
es ablished a ocused S a egic Alignmen Ma u i y, wi h IT becoming inc easingly
in ol ed.
4. Imp o ed/Managed P ocess: his ou h le el conside s IT as an inno a i e a ea,
allowing a high le el o alignmen o ein o ce he concep o IT as a alue cen e ,
which in u n enables he achie emen o compe i i e ad an age.
5. Op imized P ocess: he las le el exhibi s an op imized and ully in eg a ed IT-
business alignmen . A his le el, he en e p ise is capable o le e aging IT o
s a egic ad an age.
Ensu ing and main aining s a egic alignmen equi es a dedica ed e o o ampli y he ac o s
ha acili a e alignmen (enable s) and, a he same ime, ac i ely mi iga e he ba ie s ha
p e en i (inhibi o s) (Lu man, 2001).
30
5. FRAMEWORK PROPOSAL
The implemen a ion o AI in banking ep esen s a s a egic ans o ma ion ha imp o es he
quali y and e iciency o he se ices p o ided, po en ially e olu ionizing co e banking
unc ions and i s g adual bu widesp ead adop ion aises se e al c i ical conside a ions –
add essed in p e ious sec ions – emphasizing ha a s uc u ed app oach is needed o ha ness
AI's ull po en ial e ec i ely.
A e an ex ensi e e iew o he cu en landscape o AI applica ions in inancial se ices, along
wi h he echnological ad ancemen s and he ope a ional challenges i aces, his sec ion
in oduces a comp ehensi e amewo k designed o guide banks h ough he sys ema ic
in eg a ion o AI echnologies in o hei exis ing ope a ions.
This amewo k has been designed o add ess he challenges aced by his echnological
implemen a ion, p o iding a simple p ocess o banks o pu sue in hei implemen a ion,
emphasizing no jus he echnical aspec s o AI in eg a ion, bu also he ope a ional and
human ac o s ha a e pi o al o a success ul ou come. Fu he mo e, i conside s app op ia e
communica ion and collabo a ion, ensu ing ha he AI solu ions implemen ed a e p ope ly
aligned wi h he bank's o e all s a egic goals and ope a ional equi emen s.
In essence, h ough his s uc u ed app oach, banks may manage he complexi ies o AI
inco po a ion, con e ing challenges in o oppo uni ies o inno a ion and u he expansion.
This sec ion commences wi h he Assump ions, which es ablish he undamen al ounda ions
ha se e as he basis o he subsequen analysis. Subsequen ly, he P oposal sec ion
p esen s he p oposed amewo k in de ail, desc ibing i s s uc u e and in ended applica ion.
The ea e , he E alua ion examines he applica ion o he amewo k, analyzing he
in e ac ion be ween AI echnologies and banking se ices. The chap e concludes wi h a
demons a ion o a use case and discussion o he amewo k, in which he esul s a e
subjec ed o igo ous examina ion, alida ed agains he eal-wo ld use case, and discussed in
o de o p o ide meaning ul conclusions ega ding he impac o AI on he banking landscape.
5.1. ASSUMPTIONS
Based on he ex ensi e e iew o exis ing li e a u e on he in e sec ion be ween AI and
banking se ices, coupled wi h he unde s anding o he a ailable echnologies and hei
adop ion wi hin he indus y, i was es ablished ha a s a egy o iden i y and ad ance he
implemen a ion o AI echnologies in a chosen banking p ocess should conside he ollowing
poin s:
▪ Banking depa men s can be sys ema ically classi ied using a s uc u ed able,
simila o he APQC's ecognized P ocess Classi ica ion F amewo k (PCF) (APQC &
IBM, 2020).
37
5.3.1. Iden i ica ion o needs and objec i es
I is assumed ha he Cus ome Rela ionship Managemen (CRM) eam in he Ma ke ing
depa men was selec ed o in e en ion and ha insigh s in o i s ope a ions we e ga he ed
du ing a mee ing wi h he depa men manage :
▪ Wha a e you depa men 's p ima y esponsibili ies and objec i es?
▪ “The CRM eam is p ima ily esponsible o managing clien s' ela ionships,
unde s anding hei needs, and imp o ing clien e en ion s a egies and
ou objec i es include inc easing he e en ion o cus ome s, imp o ing
hei sa is ac ion a es, and au oma ing cus ome segmen a ion p ocesses
o enable mo e pe sonalized ma ke ing campaigns.”
▪ Wha a e he bigges challenges acing he depa men in i s ope a ions?
▪ “One o he depa men 's p incipal challenges is o enhance cus ome
engagemen in he con ex o in ensi ying compe i ion and e ol ing
cus ome expec a ions.”
▪ Lis he asks ha consume he mos ime in you daily wo k ou ine.
▪ “The mos ime-consuming asks a e cu en ly da a analysis and epo ing,
along wi h cus ome segmen a ion o a ge ed campaigns.”
▪ How would you p io i ize he needs o AI in eg a ion in you depa men ?
▪ “AI in eg a ion would be p io i ized in a eas ha could immedia ely
imp o e ope a ional e iciency and cus ome sa is ac ion.”
Fu he mo e, so chu n (cus ome segmen a ion/ e en ion) was p io i ized as a c i ical issue
o imp o e in his con ex , since i has a balanced ela ionship be ween he wo me ics: he
possible e u n alue is ca ego ized as ‘Ve y High’ since i has a di ec impac on mone a y
alue by e aining e enue ha would o he wise be los and inc eases cus ome sa is ac ion
by p o iding mo e pe sonalized se ices, and he bank's exis ing AI capabili ies and cloud-
based in as uc u e p o ide a solid ounda ion, esul ing in a mode a e e o equi ed.
Table 5.3.1 - P io i y Sco e o So Chu n.
Acco ding o his s age o he amewo k, an adequa e AI echnology o he se ice mus be
chosen. The chosen ca ego y o his issue, wi h he aid o he in o ma ion p esen ed in Annex
I, was Ma ke and Sell P oduc s and Se ices (10004) - and, wi h he suppo o he lowcha ,
since ‘P edic ion’ was he ype o AI conside ed he mos adequa e and we ha e labeled da a,
he ollowing AI echnologies ha e been iden i ied as sui able o add essing so chu n:
▪ Supe ised Lea ning (Classi ica ion): Classi ica ion models may accu a ely p edic
he p obabili y o so chu n by analyzing labeled his o ical da a.

38
▪ Deep Lea ning (Neu al Ne wo ks): Deep lea ning neu al ne wo ks can cap u e
complex pa e ns and ela ionships in la ge da ase s, p o iding mo e p ecise
p edic ions and deepe insigh s in o ac o s con ibu ing o so chu n.
▪ Time Se ies Fo ecas ing: Time se ies o ecas ing models can iden i y ends and
seasonal pa e ns in cus ome beha io o e ime, allowing o ea ly de ec ion o
po en ial chu n and imely in e en ion s a egies.
5.3.2. E alua ion and Selec ion o AI Technologies
The equi emen s ha any p oposed AI solu ion mus mee , conside ing he objec i es se ,
we e documen ed:
▪ Budge : he budge needs o conside he cos s associa ed wi h acqui ing o
de eloping an AI solu ion. Gi en he compe i i e ma ke , a mode a e o high
in es men is expec ed.
▪ Da a managemen : since cloud se ices a e al eady in ope a ion, he in es men in
da a managemen could be mo e e icien and less cos ly.
▪ In eg a ion and ope a ional cos s: ope a ing cos s may be educed due o exis ing
AI implemen a ions and he in eg a ion e o migh ocus on compa ibili y and
expansion, wi h a mode a e budge commi men .
▪ So wa e Compa ibili y: he cu en so wa e is p obably compa ible wi h
ad anced AI solu ions, so he so wa e upg ade e o should be ela i ely low.
▪ Secu i y: inc emen al upda es o secu i y measu es a e equi ed wi h mode a e
e o due o p e ious AI implemen a ions.
▪ Main enance and Suppo : ou ine main enance p ac ices ha a e al eady in place
should s eamline he inco po a ion o new AI echnologies, equi ing a mode a e
le el o addi ional e o .
The assessmen o he impac s o each echnology chosen helps in making a s a egic decision
abou which one migh be bes sui ed o add essing so chu n e ec i ely wi hin he bank.
Addi ionally, he able men ioned in S ep 2 has been comp ehensi ely illed o acili a e his
decision-making p ocess:
Table 5.3.2 - E alua ion o AI echnologies o add essing So Chu n.
Rega ding he educ ion o p ocess complexi y, he h ee echnologies a e highly e ec i e due
o hei abili y o au oma e and op imize decision-making p ocesses. In addi ion, Time Se ies
Fo ecas ing is also aluable o p edic ing ends and iden i ying ea ly signs o po en ial so
chu n, p o iding c ucial insigh s ha may help p e-emp cus ome disengagemen .
39
The ope a ional cos s associa ed wi h Supe ised Lea ning and Time Se ies Fo ecas ing a e
conside ed mode a e due o he bank's al eady implemen ed cloud se ices, which p o ide
scalable and cos -e ec i e compu ing esou ces. In con as , Deep Lea ning incu s highe
ope a ional cos s due o he possible employee aining equi ed o e ec i ely deploy and use
hese echnologies, as well as he complexi y and sophis ica ion o he models.
Fu he mo e, Time Se ies Fo ecas ing has mode a e po en ial e enue imp o emen by
p edic ing ends and seasonal pa e ns. In compa ison, Supe ised Lea ning and Deep
Lea ning ha e highe po en ial due o hei ad anced p edic ion capabili ies, which can mo e
accu a ely iden i y po en ial chu n and op imize cus ome e en ion s a egies by analyzing
complex pa e ns and ela ionships in as amoun s o da a. Addi ionally, he e a e inhe en
isks associa ed wi h hese echnologies, including po en ial biases in models and da a p i acy
conce ns, which equi e igo ous isk managemen .
Finally, he in eg a ion o Deep Lea ning is complica ed by he scale and complexi y o he
da a sys ems in ol ed. Al hough cloud in as uc u e p o ides scalable esou ces and high
compu a ional powe , he na u e o Deep Lea ning s ill equi es ca e ul planning and
specialized knowledge o e ec i e deploymen . Time Se ies Fo ecas ing, on he o he hand,
is ypically easie o in eg a e due o i s simple model s uc u e, while Supe ised Lea ning
p esen s some challenges, bu i is s ill e y use - iendly.
A e an exhaus i e analysis, Supe ised Lea ning was selec ed o deal wi h so chu n,
p esen ing i sel as a iable and e ec i e solu ion.
5.3.3. Design and In eg a ion + Implemen a ion and Moni o ing
Wi h e e ence o he able desc ibed in he amewo k, he hi d phase o he
implemen a ion p ocess was exhaus i ely planned and execu ed, as desc ibed below:
▪ This p ojec was expec ed o be scheduled o comple ion o e six mon hs, s a ing
wi h a wo-mon h phase o da a p epa a ion and ini ial model aining. Following
his, one mon h o model e inemen and alida ion, and inally, a h ee-mon h
phase o in eg a ion and es ing. The schedule was espec ed. This p ojec was
implemen ed in he could in as uc u e al eady implemen ed wi hin he bank,
wi h a dedica ed eam o da a scien is s and da a enginee s and a complex se o
da a alloca ed in he cloud se ice.
▪ The model was designed o comply wi h he GDPR and o he ele an da a
p o ec ion egula ions, ensu ing ha all cus ome da a is handled secu ely.
▪ The in o ma ion abo e was deli e ed o he s akeholde s h ough epo s,
mee ings, and p esen a ions.
This o ganized s uc u e ensu ed a smoo h and o de ly deli e y o he p ojec . No addi ional
aining was equi ed, making he implemen a ion and moni o ing phase s aigh o wa d.
Pe o mance was e alua ed mon hly, including i s p ecision and o he key me ics.
40
5.4. EVALUATION AND DISCUSSION
To assess he ele ance o he p oposed amewo k designed o acili a e he in eg a ion o AI
in he banking indus y, in e iews we e conduc ed. These in e iews aimed o e alua e he
e ec i eness o he amewo k and o p o ide insigh in o possible imp o emen s.
To achie e his, wo dis inc p o iles we e in e iewed based on hei le el o knowledge and
expe ise: one yea ’s expe ience in he ield and a senio da a scien is .
To acili a e he e alua ion, a comp ehensi e p esen a ion, a ached in APPENDIX C, was
c ea ed and sha ed wi h he in e iewees o p o ide a clea o e iew o he amewo k and
ga he de ailed eedback.
Following his, a s uc u ed su ey was conduc ed:
▪ Do you hink ha he p oposed amewo k is use ul? Why/ why no ?
▪ Da a scien is T ainee: “Yes, i seems use ul. The s ep-by-s ep layou helps
unde s and whe e he p ocess is a and wha needs o be done nex .”
▪ Senio Da a Scien is : “Yes, I belie e he amewo k is highly bene icial.
Howe e , i appea s o be pa icula ly ad an ageous o adi ional banks
ha a e jus beginning o explo e he po en ial o AI. Besides ha , i s
s eng h comes om i s abili y o align wi h he speci ic needs o he bank,
a he han simply ollowing he la es ends, he eby p o iding a
comp ehensi e assessmen o he equi ed e o , associa ed isks, and
po en ial bene i s! This well-s uc u ed app oach is a signi ican ad an age,
se ing he ounda ion o success ul ech in eg a ion!”
▪ Do you conside he amewo k o be clea and unde s andable?
▪ Da a scien is T ainee: “Yes, I do. I eally app ecia e he lowcha ; i seems
o be an indispensable isual elemen in his con ex . I is e y easy o
unde s and and manage. Howe e , he e a e a eas wi hin i ha could
bene i om g ea e speci ici y and mo e de ailed explana ions. I would
also inco po a e anno a ions be ween ce ain s eps.”
▪ Senio Da a Scien is : “Yes, I do ind i use ul. I belie e ha i s success
doesn' ely on he use 's le el o expe ise because i uses clea ocabula y
and concep s ha a e easy o unde s and. I is well-g ounded in basic
banking p inciples, which makes i s aims and me hods clea . I also eminds
me o an ML pipeline, al hough i is mo e o a p elimina y heo e ical
amewo k!”
▪ Wha imp o emen s would you sugges o his amewo k?
▪ Da a scien is T ainee: “Beyond wha I al eady sugges ed, I belie e
in oducing eedback loops, ei he di ec ly wi hin he lowcha o as pa
o he depa men al su ey, would help in con inuous e inemen . The las
s eps o he amewo k should also ha e some isual ep esen a ion, I
hink. I also conside ha adding a componen ela ed o echnological
de elopmen s migh p e en ab up changes.”
41
▪ Senio Da a Scien is : “I unde s and he objec i e o he amewo k and
ecognize ha modi ying e en a single s ep o an auxilia y ool could be
complex due o hei in e dependence. Howe e , I belie e he bank could
bene i mo e i , o example, he de elopmen o an in eg a ion plan was
es ablished ea lie in he p ocess, no jus in he hi d s ep. The amewo k's
‘cyclical’ na u e, which I app ecia e, seems o delay eaching conclusions: i
I selec a banking p ocess and hen choose speci ic echnologies o assess
indi idually, I mus p og ess o a leas he hi d s ep o de e mine i
in eg a ion is easible be o e conside ing an al e na i e, since he imeline,
esou ces, and he equi emen s any AI solu ion mus mee a e only
de ailed in la e s eps. This app oach would ake he bank's echnological
ma u i y and capaci y in o accoun igh om he s a , p e en ing he
p oposal o un easible solu ions."
Table 5.4.1 - Ad an ages and disad an ages o he amewo k.
Ad an ages
Disad an ages
Legibili y
Scalabili y
Applicabili y
Adap abili y
Cla i y
The o e all u ili y o he amewo k ecei ed posi i e alida ion om he pa icipan s.
Howe e , he eedback also highligh ed a need o a mo e concise layou , pa icula ly
emphasizing he inclusion o isual ep esen a ions o he inal s ages o he amewo k. In
addi ion, he insigh s om he ainee we e pa icula ly aluable in assessing he amewo k's
legibili y and unde s andabili y, which is undamen al o use s wi h lowe expe ience le els.
Wi h ega d o he Senio Da a Scien is , an al e na i e app oach was p oposed: he o de o
some s eps could be al e ed in o de o enhance he p ac icali y and e ec i eness o he
p ocess o selec ing he mos app op ia e AI echnology o he selec ed p ocess.
Mo eo e , a signi ican gap iden i ied in he amewo k is he lack o speci ici y du ing he
echnology selec ion phase wi hin he lowcha . Add essing his could enhance he
adap abili y o he amewo k and make i pa icula ly bene icial o banks ha a e in he ea ly
s ages o in eg a ing AI echnologies. Fu he mo e, i would lead o decisions aligned wi h
hei speci ic con ex s and echnological capabili ies, p o iding clea e guidance.
I was also obse ed ha he absence o anno a ions and examples in c i ical decision poin s
could signi ican ly a ec use comp ehension and engagemen , making his a c i ical
equi emen o u u e wo k.
In summa y, al hough he amewo k has demons a ed i s o e all e ec i eness h ough
eedback om pa icipan s, he con ibu ions ga he ed highligh essen ial aspec s o be
imp o ed in o de o maximize i s use ulness ha may ans o m he amewo k in o a mo e
42
in ui i e and e ec i e ool and acili a e a smoo he adop ion and in eg a ion o AI
echnologies ac oss he banking sec o .

43
6. CONCLUSIONS AND FUTURE WORK
This chap e summa izes he wo k de eloped wi hin he scope o his disse a ion, clea ly
de ines he signi ican esul s, and es ablishes he basis o he ans o ma i e in eg a ion o
AI in he banking sec o and u u e wo k, men ioning i s limi a ions.
The p ima y objec i e o his s udy was o es ablish a me hodical amewo k o acili a e he
in eg a ion o A i icial In elligence echnologies in o banking ope a ions, enhancing o e all
se ice deli e y and ope a ional e iciency. The amewo k was designed o accommoda e
a ious AI applica ions, wi h so chu n se ing as one illus a i e example.
The ou -s ep amewo k was me hodically de eloped, acili a ing a sys ema ic app oach o
he selec ion and implemen a ion o AI echnologies in he banking sec o . This amewo k
no only ou lined he me hodical iden i ica ion and assessmen o banks' speci ic needs bu
also highligh ed he s a egic alignmen o AI echnologies wi h hese equi emen s o
op imize banking ope a ions. I was me iculously de eloped by syn hesizing insigh s om
ex ensi e li e a u e e iews and adap ing bes p ac ices om he indus y.
The esea ch ques ion p esen ed in he In oduc ion: “How can a s uc u ed amewo k
op imize he deploymen and managemen o a i icial in elligence echnologies in he
banking sec o ?” has been add essed h ough he demons a ion o he use case e ec i eness
p esen ed in Chap e 5.3. The esul s alida e he amewo k's po en ial as a aluable ool o
banking ins i u ions seeking o le e age AI o s a egic ad an age.
6.1. FRAMEWORK LIMITATIONS
While he amewo k designed o in eg a e AI echnologies in o banking ope a ions has
demons a ed po en ial, and despi e achie ing he o e all goals o his disse a ion, he e a e
inhe en limi a ions ha should be acknowledged:
▪ Since he amewo k is based on he APQC's classi ica ion o he banking p ocess i
may limi i s adap abili y o he speci ic ope a ional nuances o each bank since i
assumes he s anda diza ion o he se ices o e ed by banks;
▪ The lowcha p esen ed in he ini ial s age o he amewo k is ela i ely gene ic
and would bene i om being subdi ided in o mo e ela able and speci ic
ca ego ies ha align closely wi h p ac ical banking ope a ions;
▪ The absence o mul iple use case demons a ions may limi he p ac ical alida ion
o he amewo k ac oss di e en scena ios and con ex s wi hin he banking
sec o .
▪ Ce ain s eps may be implemen ed oo la e, which could esul in he p ocess
aking longe han necessa y.
▪ The ield o AI echnologies is apidly e ol ing, and he amewo k mus be
con inuously upda ed o inco po a e he la es ad ancemen s and bes p ac ices
o emain ele an and e ec i e.
44
The amewo k mus be dynamically e ined and adap ed based on ongoing eedback and
e ol ing condi ions wi hin he banking sec o .
6.2. FUTURE WORK
Once his mas e 's disse a ion is ully comple ed, i may se e as a aluable a i ac o aid in
he in eg a ion and implemen a ion o AI echnologies wi hin a banking ins i u ion. To ensu e
he con inued ele ance and e ec i eness o his amewo k, nume ous imp o emen s could
be achie ed. These enhancemen s aim o expand he amewo k's applicabili y and u ili y:
▪ The amewo k should be applied in an ac ual banking en i onmen in o de o
es ablish a p ac ical ounda ion o ongoing e inemen .
▪ Include mo e cus omizable op ions in he auxilia y ables, allowing banks o ailo
AI in eg a ion s a egies o hei speci ic ope a ional con ex s and cus ome needs.
▪ Inco po a e a b oade ange o use case examples ac oss di e en banking se ices
o illus a e di e se applica ions and bene i s.
▪ Conside eo de ing he amewo k's implemen a ion and include use case
examples o compa ison wi h es ablished p ac ices.
▪ As he egula ion and go e nance o AI become inc easingly signi ican , pa icula ly
wi h he in oduc ion o he Eu opean AI Ac , aligning his amewo k wi h he
equi emen s o he AI Ac would be highly bene icial. Fu he mo e, his alignmen
would enhance he amewo k's ocus on isk managemen , anspa ency, and
e hical conside a ions, he eby making i mo e obus and sui ed o mee indus y
s anda ds and egula o y expec a ions.
By add essing hese, he amewo k could p obably e ol e in o a mo e obus ool ha mee s
he cu en demands o he banking indus y and is also adap able o u u e changes and
challenges, ensu ing i s iabili y and success in acili a ing AI in eg a ion.
45
BIBLIOGRAPHICAL REFERENCES
Abioye, S., Oyedele, L., Akanbi, L., Ajayi, A., Delgado, J., Bilal, M., Akinade, O., & Ahmed, A.
(2021). A i icial in elligence in he cons uc ion indus y: A e iew o p esen s a us,
oppo uni ies and u u e challenges. Jou nal o Building Enginee ing, 44.
h ps://doi.o g/10.1016/j.jobe.2021.103299
Ahmadi, S. (2024). A Comp ehensi e S udy on In eg a ion o Big Da a and AI in Financial
Indus y and i s E ec on P esen and Fu u e Oppo uni ies. In e na ional Jou nal o
Cu en Science Resea ch and Re iew, 7(1). h ps://doi.o g/10.47191/ijcs /V7-i1-07
Ake ka , R., & Sajja, P. (2010). Knowledge-Based Sys ems. Jones & Ba le Lea ning.
Akh e , S., Ahmad, M., Chibb, M., Zai, A., & Yaqoob, M. (2024). A i icial In elligence In The
21s Cen u y: Oppo uni ies, Risks And E hical Impe a i es. Educa ional Adminis a ion:
Theo y and P ac ice. h ps://doi.o g/10.53555/kuey. 30i5.3125
AL-Dosa i, K., Fe ais, N., & Kucuk a , M. (2022). A i icial In elligence and Cybe De ense
Sys em o Banking Indus y: A Quali a i e S udy o AI Applica ions and Challenges.
Cybe ne ics and Sys ems, 55. h ps://doi.o g/10.1080/01969722.2022.2112539
Al-Tamimi, H. (2007). Banks’ isk managemen : a compa ison s udy o UAE na ional and
o eign banks. The Jou nal o Risk Finance, 8.
h ps://doi.o g/10.1108/15265940710777333
Alzubi, J., Nayya , A., & Kuma , A. (2018). Machine Lea ning om Theo y o Algo i hms: An
O e iew. Jou nal o Physics Con e ence Se ies. h ps://doi.o g/10.1088/1742-
6596/1142/1/012012
Ame ican P oduc i i y & Quali y Cen e (APQC). (2019). Banking P ocess Classi ica ion
F amewo k e sion 7.2.1.
APQC, & IBM. (2020). Banking P ocess Classi ica ion F amewo k.
h ps://www.apqc.o g/ esou ce-lib a y/ esou ce-lis ing/apqc-p ocess-classi ica ion-
amewo k-pc -banking-pc -pd -1
A ie a, A., Díaz-Rod íguez, N., Del Se , J., Benne o , A., Tabik, S., Ba bado, A., Ga cia, S., Gil-
Lopez, S., Molina, D., Benjamins, R., Cha ila, R., & He e a, F. (2019). Explainable
A i icial In elligence (XAI): Concep s, axonomies, oppo uni ies and challenges owa d
esponsible AI. h ps://doi.o g/10.1016/j.in us.2019.12.012
Biswal, K. C. (2015). Eme ging T ends in he Indian Banking Sec o - Challenges &
Oppo uni ies. In e na ional Jou nal o Ad ances in A s, Sciences and Enginee ing, 3.
h ps://joill.in/ijoaase.com/uploads/3/2/1/7/3217988/eme ging_ ends_in_ he_indian
_banking_sec o _challenges__oppo uni ies.pd
46
Bo y, P. (2019). Deep new: The shi ing na a i es o a i icial in elligence om Deep Blue o
AlphaGo. Con e gence The In e na ional Jou nal o Resea ch in o New Media
Technologies. h ps://doi.o g/10.1177/1354856519829679
B igh wood, S. (2024). Challenges and oppo uni ies in adop ing AI in he inance indus y
Au ho .
h ps://www. esea chga e.ne /publica ion/379078913_Challenges_and_oppo uni ies_
in_adop ing_AI_in_ he_ inance_indus y_Au ho
B oby, D. (2021). Financial echnology and he u u e o banking. Financial Inno a ion.
h ps://doi.o g/10.1186/s40854-021-00264-y
B ocke, J., He ne , A., & Maedche, A. (2020). In oduc ion o Design Science Resea ch.
Resea chGa e. h ps://doi.o g/10.1007/978-3-030-46781-4_1
Ca è, R. (2018). Sus ainable Banking: Issues and Challenges. Sp inge .
Cebi, F., Aydin, O. F., & Gozlu, S. (2010). Bene i s o Knowledge Managemen in Banking.
Jou nal o T ansna ional Managemen . h ps://doi.o g/10.1080/15475778.2010.525486
Chen, L. (2010). Business–IT alignmen ma u i y o companies in China. In o ma ion &
Managemen . h ps://doi.o g/10.1016/j.im.2009.09.003
Coa es, J. (1974). Some Me hods and Techniques o Comp ehensi e Impac Assessmen .
Technological Fo ecas ing and Social Change. h ps://doi.o g/10.1016/0040-
1625(74)90035-3
Daniel, S., & Vi ky, I. (2024). Technological Fu u es, Oppo uni ies and Di icul ies in A i icial
In elligence and Machine Lea ning.
h ps://www. esea chga e.ne /publica ion/377150634_Technological_Fu u es_Oppo
uni ies_and_Di icul ies_in_A i icial_In elligence_and_Machine_Lea ning
De Haas, R. (2023). Sus ainable Banking. h p://dx.doi.o g/10.2139/ss n.4620166
Di ya, D. (2024). A i icial In elligence In Finance. Inno a e & T ans o m Re olu ionizing
Business wi h B eak h ough C ea i i y.
h ps://www. esea chga e.ne /publica ion/380266484_ARTIFICIAL_INTELLIGENCE_IN_
FINANCE
Donepudi, P. (2017). Machine Lea ning and A i icial In elligence in Banking. Enginee ing
In e na ional . h ps://doi.o g/10.18034/ei. 5i2.490
Dumasiya, R. (2023). A S udy o Impac , Issues and Challenges o Recen Public and P i a e
Sec o Bank Me ge on S akeholde s. In e na ional Mul idisciplina y Resea ch Jou nal,
9.
53
APPENDIX A

54
APPENDIX B
55
APPENDIX C
In o ma on
Managemen
School
Ins u o Supe io de Es a s ca e Ges o da In o ma o
Uni e sidade o a de Lisboa
Ac edi a es e Ce ca es
Mas e s Deg ee P og am in
Da a Science and Ad anced Analy cs
Ana Ri a Figuei edo Ta a es
Supe iso : P o . D . V o San os
Lack o Comp ehensi e F amewo ks in he
banking indus y.
Rapid Technological E olu on.
Po en al o enhance e ciency, pe sonaliza on
and cus ome sa s ac on.
De elopmen o a
amewo k ha may be
applied by he banking
sec o o in
i s se ices.
The amewo k aims o
o
AI echnologies
The p ima y goal o he
selec ed echnologies is
o enhance he o
banking se ices and
boos cus ome
.
56
Iden ca ono
needs and
objec es
Selec ono AI
echnologies Design and
In eg a on Implemen a on
and Moni o ing
Iden ca ono
needs and
objec es
Su ey
Selec he
in e ened
banking
depa men .
Collec insigh s
om he
selec ed
depa men .
Selec he
p ocesses ha
need o be
add essed o
op mized.
So he
p ocesses
acco ding o
hei p io i y.
Selec adequa e
AI Technologies.
Iden ca ono
needs and
objec es
Su ey
Wha a e you depa men 's p ima y and ?
Wha a e he acing he depa men in i s
ope a ons?
Lis he asks ha consume he in you daily wo k ou ne.
How would you he needs o AI in eg a on in you
depa men ?
57
Selec he
in e ened
banking
depa men .
Collec insigh s
om he
selec ed
depa men .
Selec he
p ocesses ha
need o be
add essed o
op mized.
So he
p ocesses
acco ding o
hei p io i y.
Selec adequa e
AI Technologies.
Iden ca ono
needs and
objec es
Flowcha
Flowcha
58
Selec on o AI
echnologies
Documen he
equi emen s ha
any p oposed AI
solu on mus
mee .
Map he chosen
echnologies o
he selec ed
banking p ocess
Selec he mos
p omising
echnology o he
objec e.
Budge
Da a Managemen
In eg a on and Ope a onal Cos s
So wa e compa bili y
Secu i y

59
Design and
In eg a on
De elop an
in eg a on
plan.
Upda e he
able wi h any
upda es ela ed
o compliance
o isk
mi ga on
measu es.
Communica e
deploymen
schedules and
expec a ons o
all ele an
s akeholde s.
T aining and
suppo
(op onal).
Includes Timelines and Resou ces
Implemen a on
and Moni o ing
T ansi he AI
sys em in o
ope a onal use.
Pe o mance
Moni o ing.
Regula ly
documen and
epo .
Ini a e
con nuous
imp o emen
cycles
Insigh s om moni o ing and eedback om use s and s akeholde s a e used o e ne
and enhance AI sys em pe o mance.
Ins u o Supe io de Es a s ca e Ges o da In o ma o
Uni e sidade o a de Lisboa
Add ess: Campus de Campolide, 1070 12 Lisboa, Po ugal
Phone: 1 21 2 10 Fax: 1 21 2 11
Ac edi a es e Ce ca es
60
ANNEXES
ANNEX I – APQC’s P ocess F amewo k, Banking Indus y. Adap ed by he au ho .
61
62
69

70
71