37
In e na ional Jou nal o Ad ance and Applied Resea ch
www.ijaa .co.in
ISSN – 2347-7075
Impac Fac o – 8.141
Pee Re iewed
Bi-Mon hly
Vol. 6 No. 38
Sep embe - Oc obe - 2025
A i icial In elligence in Educa ion: A Comp ehensi e Re iew o Resea ch
T ends, E hical Conside a ions, and Technological Ad ancemen s
D . Manjusha Ko hawade1 & D . Minal Bhosale2
1,2D . D. Y. Pa il A s, Comme ce & Science College, Aku di, Pune-44
Co esponding Au ho – D . Manjusha Ko hawade
DOI - 10.5281/zenodo.17309861
Abs ac :
A i icial In elligence (AI) is inc easingly eshaping educa ional p ac ices, o e ing enhanced
pe sonaliza ion, au oma ed assessmen , and no el eaching modali ies. This e iew syn hesizes ecen
ad ancemen s in AI applied o educa ion (AIED) h ough an analysis o key li e a u e, emphasizing
esea ch ends, e hical challenges, human-cen e ed AI design, and echnological inno a ions such as
social obo ics and adap i e sys ems. D awing upon ecen sys ema ic s udies and heo e ical
discussions, his pape p o ides a comp ehensi e unde s anding o AIED’s concep ual landscape,
iden i ies p essing esea ch gaps, and sugges s u u e di ec ions o e hical, e ec i e, and inclusi e
adop ion in di e se educa ional con ex s.
Keywo ds: A i icial In elligence, Educa ion, E hics, Pe sonalized Lea ning, In elligen Tu o ing
Sys ems, Social Robo s, Collabo a i e Lea ning
In oduc ion:
A i icial in elligence (AI) has
eme ged as a ans o ma i e o ce in
educa ion, in luencing eaching me hods,
lea ning en i onmen s, and adminis a i e
p ocesses (Age alk and Ka lsson, 2020). AI in
educa ion (AIED) sys ems le e age
compu a ional models o mimic cogni i e asks
such as p oblem-sol ing, decision-making, and
language unde s anding, enabling adap i e
lea ning, au oma ed g ading, and da a-d i en
educa ional analy ics. These sys ems can
p o ide pe sonalized ins uc ion ailo ed o
indi idual lea ne ’s needs, op imize
educa ional esou ces, and p omo e g ea e
accessibili y and engagemen (Chen e al.,
2020; Boulay, 2023).
The apid g ow h o AI echnologies,
coupled wi h digi aliza ion accele a ed by
ecen global e en s like he COVID-19
pandemic, has spo ligh ed bo h oppo uni ies
and challenges associa ed wi h AIED. While
AI p omises ans o ma i e imp o emen s
ac oss educa ional le els—K-12, highe
educa ion, and li elong lea ning—conce ns
a ound e hics, equi y, and human agency ha e
become inc easingly salien (Akgun and
G eenhow, 2022).
Building on he ex ensi e body o
esea ch eme ging in he las decade, his pape
sys ema ically e iews key ends and deba es
in AIED. We analyze ounda ional and
con empo a y s udies, elucida e he main AI
applica ions and hei pedagogical
implica ions, and highligh c i ical e hical and
social issues. The pape is s uc u ed o
add ess:
(i) he p ime AI echnologies shaping
educa ion;
(ii) human-cen e ed design
app oaches;
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
D . Manjusha Ko hawade & D . Minal Bhosale
38
(iii) eme ging e hical challenges in K-
12 and beyond;
(i ) he ole o AI-powe ed social
obo s and collabo a i e pla o ms; and
( ) u u e esea ch di ec ions o bols e
equi able, e ec i e AI in eg a ion.
Li e a u e Re iew:
1. E olu ion and Scope o AI in Educa ion
Resea ch:
The incep ion o AI in educa ion
esea ch ocused p ima ily on in elligen
u o ing sys ems (ITS) and compu e -assis ed
ins uc ion, aimed a indi idualizing lea ning
in con olled en i onmen s (Age alk and
Ka lsson, 2020). O e ime, ad ances in
machine lea ning, na u al language p ocessing,
and senso echnologies expanded p ospec s
o adap i e, in e ac i e, and collabo a i e
educa ional AI sys ems (Chen e al., 2020).
Chen e al. (2020) highligh ha
con empo a y AIED emb aces da a-d i en,
gene a i e AI models ha no only espond o
bu an icipa e lea ne needs by in eg a ing
mul imodal da a, including a ec i e and
beha io al in o ma ion. Guan e al. (2020)
emphasize he explosion o AI esea ch ou pu
o e he pas wo decades, wi h ends poin ing
owa d inc easingly human-cen ic AI models
designed o in eg a e seamlessly wi hin
exis ing pedagogical amewo ks.
Concu en o echnological ad ances,
e hical and social dimensions ha e gained
c i ical a en ion. Akgun and G eenhow (2022)
a gue ha add essing bias, p i acy,
accoun abili y, and anspa ency is i al,
especially in o ma i e K-12 con ex s whe e
de elopmen al sensi i i ies a e pa amoun .
2. P incipal AI Applica ions in Educa ion:
Pe sonalized Lea ning and In elligen
Tu o ing:
Pe sonalized lea ning, unde pinned by
adap i e algo i hms, is widely s udied o i s
abili y o ailo con en , complexi y, and
modali y pe lea ne 's cogni i e and emo ional
p o ile. AI-based ITS simula e human u o s by
p o iding cus omized explana ions, p ac ice
p oblems, and eal- ime eedback (Chen e al.,
2020; Maghsudi e al., 2021). Bicknell e al.
(2023) explain how pla o ms like Duolingo
use AI o indi idualize language lea ning,
enhancing e en ion and mo i a ion.
Au oma ed Assessmen and Feedback:
AI-powe ed sys ems au oma e he
e alua ion o open- ex , coding, and mul iple-
choice assessmen s, o e ing consis en sco ing
and immedia e, o ma i e eedback (Gonzalez-
Cala ayud e al., 2021). These sys ems elie e
educa o s’ adminis a i e bu den, allowing
mo e ime o ins uc ional design and lea ne
engagemen (Boulay, 2023).
Social Robo s:
Belpaeme and Tanaka (2022) discuss
he bu geoning ole o social obo s as
educa o s and companions, emphasizing hei
abili y o engage young lea ne s h ough social
p esence, mul imodal communica ion, and
emo ional a unemen . These obo s ha e
demons a ed e icacy in language
de elopmen , social skills aining, and
inclusi e educa ion o ma ginalized
popula ions.
Collabo a i e AI Technologies:
Ande sen e al. (2022) desc ibe AI-
in eg a ed collabo a i e lea ning en i onmen s
whe e block-based p og amming acili a es
pee in e ac ion and sca olded p oblem-
sol ing. Such human-cen e ed designs
p io i ize AI as assis i e a he han di ec i e,
suppo ing social cons uc i is lea ning.
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
D . Manjusha Ko hawade & D . Minal Bhosale
39
3. Bibliome ic Analysis:
Bibliome ic analysis p o ides a
quan i a i e lens o examine he e olu ion and
hema ic s uc u e o AIED esea ch. Ka i ha
(2024) analyzed 775 pee - e iewed Scopus-
indexed publica ions om 2000 o 2022,
ma king a signi ican ise in annual ou pu s,
g owing om minimal ea ly con ibu ions o
o e 300 pape s annually in ecen yea s.
China and he USA lead global con ibu ions,
alongside subs an ial mul ina ional
collabo a ions, unde sco ing a ib an , global
esea ch communi y.
Keywo d co-occu ence mapping
e eals majo esea ch clus e s a ound
adap i e lea ning, in elligen u o ing, neu al
ne wo ks, and educa ional da a mining,
indica ing dual emphases on echnical
me hodologies and pedagogical applica ions
(Ka i ha, 2024). The li e a u e spans di e se
jou nals specializing in educa ional
echnology, compu e science, and
in e disciplina y in o ma ion sys ems.
Addi ional bibliome ic s udies
highligh eme ging in e es in AI cha bo s o
language lea ning and o ma i e assessmen s,
e lec ing e ol ing p io i ies owa ds
con e sa ional AI and use expe ience (Liu e
al., 2024). This expansi e bibliome ic
mapping si ua es AIED esea ch as bo h
ma u e and apidly e ol ing, wi h in ensi ying
ocus on human-cen e ed and e hics-awa e AI
design.
4. Con en Analysis:
Con en analysis complemen s
bibliome ic indings by explo ing hema ic
dep h wi hin sampled li e a u es spanning
di e se AI applica ions and socio-e hical
dimensions. Co e applica ions iden i ied
include adap i e u o ing sys ems deli e ing
eal- ime pe sonalized eedback, au oma ed
g ading enhancing o ma i e assessmen
quali y, and embodied social obo s acili a ing
engagemen and social lea ning (Chen e al.,
2020; Belpaeme and Tanaka, 2022).
E hical conce ns pe ade he con en ,
ocusing on da a p i acy, algo i hmic ai ness,
and socio- echnical inclusi i y, wi h calls o
pa icipa o y AI co-design in ol ing educa o s
and lea ne s (Akgun and G eenhow, 2022).
Teache s’ e ol ing oles amids AI in eg a ion
and he need o p o essional de elopmen a e
ecu en hemes, emphasizing he syne gy
a he han subs i u ion o human and machine.
Implemen a ion challenges including
con ex ual adap abili y, in as uc u al
dispa i ies, and sus ained e alua ion a e
e iden , unde sco ing he complexi y o eal-
wo ld AI deploymen . These quali a i e
insigh s illumina e pa hways o designing
anspa en , equi able, and pedagogically
aligned AI educa ional ecosys ems (Chiu e
al., 2023).
5. Cu en T ends in AI in Educa ion
(2024–2025):
Recen analyses documen
accele a ing adop ion o gene a i e AI, i ual
u o s, AI eaching assis an s, and gami ied
educa ional echnologies (Sp ings, 2025;
EIMT, 2025). These AI capabili ies gene a e
ich educa ional con en , adap lea ning
ajec o ies, p o ide 24/7 academic suppo ia
cha bo s, and enhance inclusion h ough
mul imodal accessibili y ea u es like speech-
o- ex and emo ion ecogni ion.
Sp ings (2025) ou lines key ends including:
Gene a i e AI’s ole in con en c ea ion,
deli e ing ailo ed lessons, exe cises,
and mul imedia lea ning objec s wi h
ime and cos e iciencies.
AI-powe ed s udy companions and
coaches o e ing pe sonalized eedback,
goal se ing, and con ex ualized
assis ance.
G owing in eg a ion o AI agen s in
Lea ning Managemen Sys ems (LMS)
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
D . Manjusha Ko hawade & D . Minal Bhosale
40
o au oma e g ading, a endance, and
esou ce alloca ion while enhancing
s uden engagemen .
Inclusi e echnologies suppo ing
lea ne s wi h disabili ies and a ied
lea ning p e e ences.
Fu he mo e, EIMT (2025) epo s a
widening up ake o AI u o ing sys ems ac oss
educa ion sec o s, a ise in au oma ed g ading,
and expansion o AI-d i en analy ics o ea ly
in e en ion and ca ee guidance.
6. E hical Challenges and Conside a ions:
Akgun and G eenhow (2022)
ex ensi ely discuss e hical challenges inhe en
in K-12 deploymen s, s essing he impo ance
o balancing inno a ion wi h sa egua ding
s uden p i acy, p e en ing algo i hmic biases,
and main aining anspa ency in AI decision-
making. Issues o da a secu i y, consen , and
minimiza ion o ha m a e c i ical, especially
when AI sys ems collec sensi i e beha io al
o biome ic da a.
Boulay (2023) con ex ualizes e hical
conce ns wi hin b oade socio-poli ical
amewo ks, ad oca ing o pa icipa o y
design app oaches in ol ing educa o s,
s uden s, and communi ies o co-c ea e AI
sys ems ha e lec di e se alues and needs.
Teache oles also e ol e amid au oma ion;
Celik e al. (2022) highligh he need o
eache p o essional de elopmen o in eg a e
AI e ec i ely wi hou unde mining
p o essional au onomy o pedagogical
expe ise.
Discussion:
The in eg a ion o a i icial
in elligence in o educa ion he alds a
ans o ma i e e a, o e ing unp eceden ed
oppo uni ies o ailo lea ning expe iences,
s eamline educa ional adminis a ion, and
ex end inclusi e educa ional access. Howe e ,
he jux aposi ion o hese oppo uni ies wi h
mul i ace ed challenges equi es ca e ul
sc u iny and s a egic ac ion o ensu e
esponsible, equi able, and e ec i e AI
adop ion in educa ion sys ems wo ldwide.
Implemen a ion challenges include da a
p i acy isks, algo i hmic biases, digi al
di ides, and echnology accep ance ba ie s
among educa o s and s uden s (TeachBe e .ai,
2025; Bhaska , 2024). The balance be ween
AI-d i en au oma ion and essen ial human
in e ac ion emains delica e, as educa ion’s
social and emo ional dimensions esis ull
mechaniza ion (Celik e al., 2022).
E hical go e nance mus p io i ize
anspa ency, accoun abili y, and pa icipa o y
design o main ain s akeholde us and
add ess ai ness and p i acy conce ns (Boulay,
2023).
S a egic p io i ies in ol e os e ing AI
li e acy among educa o s, in es ing in
equi able in as uc u e, and designing human-
cen e ed AI ools ha suppo collabo a i e
and inclusi e pedagogy (Ande sen e al., 2022;
Akgun and G eenhow, 2022).
Eme ging oppo uni ies in social obo ics,
gene a i e AI, and collabo a i e pla o ms
p omise en iched, accessible lea ning
amewo ks, p o ided hey a e de eloped wi h
ongoing e hical igilance and pedagogical
alignmen (Belpaeme and Tanaka, 2022;
Bicknell e al., 2023).
Conclusion:
AI in educa ion is a mul i ace ed,
apidly e ol ing domain shaping how lea ning
and ins uc ion a e conduc ed, p omising
g ea e pe sonaliza ion, e iciency, and
inclusi i y. This e iew syn hesizes ad ances
ac oss pe sonalized u o ing, au oma ed
assessmen s, social obo ics, and human-
cen e ed collabo a i e sys ems, highligh ing
bo h echnological possibili ies and e hical
impe a i es.
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
D . Manjusha Ko hawade & D . Minal Bhosale
41
Responsible and equi able AI adop ion
necessi a es ongoing dialogue among
echnologis s, educa o s, policymake s, and
lea ne s o ensu e AI en iches educa ion
wi hou comp omising human alues. Fu u e
esea ch ex ending beyond echnical e icacy
owa d social jus ice and cul u al
esponsi eness will unlock AI’s ans o ma i e
po en ial in global educa ion.
Re e ences:
1. Age alk, P. J., & Ka lsson, F. (2020).
A e ac ual and empi ical
con ibu ions in in o ma ion sys ems
esea ch. Eu opean Jou nal o
In o ma ion Sys ems, 29, 109-113.
h ps://doi.o g/10.1080/0960085X.202
0.1743051
2. Akgun, S., & G eenhow, C. (2022).
A i icial in elligence in educa ion:
Add essing e hical challenges in K-12
se ings. AI and E hics, 2, 431-440.
h ps://doi.o g/10.1007/s43681-021-
00096-7
3. Ande sen, R., Mø ch, A. I., &
Li he land, K. T. (2022). Collabo a i e
lea ning wi h block-based
p og amming: In es iga ing human-
cen e ed a i icial in elligence in
educa ion. Beha iou & In o ma ion
Technology, 41, 1830-1847.
h ps://doi.o g/10.1080/0144929X.202
2.2083981
4. Belpaeme, T., & Tanaka, F. (2022).
Social Robo s as Educa o s. OECD.
h ps://www.oecd-
ilib a y.o g/si es/1c3b1d56-
en/index.h ml?i emId=/con en /compo
nen /1c3b1d56-en
5. Bicknell, K., B us , C., & Se les, B.
(2023). How Duolingo’s AI Lea ns
Wha You Need o Lea n. IEEE
Spec um.
h ps://spec um.ieee.o g/duolingo
6. Boulay, B. D. (2023). A i icial
In elligence in Educa ion and E hics.
In O. Zawacki-Rich e & I. Jung
(Eds.), Handbook o Open, Dis ance
and Digi al Educa ion (pp. 93-108).
Sp inge Singapo e.
h ps://doi.o g/10.1007/978-981-19-
2080-6_6
7. Chen, L., Chen, P., & Lin, Z. (2020).
A i icial in elligence in educa ion: A
e iew. IEEE Access, 8, 75264-75278.
h ps://doi.o g/10.1109/ACCESS.2020
.2988510
8. Chiu, T. K. F., Xia, Q., Zhou, X.,
Chai, C. S., & Cheng, M. (2023).
Sys ema ic li e a u e e iew on
oppo uni ies, challenges, and u u e
esea ch ecommenda ions o a i icial
in elligence in educa ion. Compu e s
and Educa ion: A i icial In elligence,
4, 100118.
h ps://doi.o g/10.1016/j.caeai.2022.10
0118
9. Celik, I., Dinda , M., Muukkonen, H.,
& Jä elä, S. (2022). The p omises
and challenges o a i icial in elligence
o eache s: A sys ema ic e iew o
esea ch. TechT ends, 66, 616-630.
h ps://doi.o g/10.1007/s11528-022-
00715-y
10. C omp on, H., Jones, M. V., & Bu ke,
D. (2022). A o dances and challenges
o a i icial in elligence in K-12
educa ion: A sys ema ic e iew.
Jou nal o Resea ch on Technology in
Educa ion, 1-21.
h ps://doi.o g/10.1080/15391523.202
2.2121344
11. Guan, C., Mou, J., & Jiang, Z. (2020).
A i icial in elligence inno a ion in
educa ion: A wen y-yea da a-d i en
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
D . Manjusha Ko hawade & D . Minal Bhosale
42
his o ical analysis. In e na ional
Jou nal o Inno a ion S udies, 4, 134-
147.
h ps://doi.o g/10.1016/j.ijis.2020.09.0
01
12. Maghsudi, S., Lan, A., Xu, J., & an
De Schaa , M. (2021). Pe sonalized
educa ion in he a i icial in elligence
e a: Wha o expec nex . IEEE Signal
P ocessing Magazine, 38, 37-50.
h ps://doi.o g/10.1109/Msp.2021.305
5032
13. Sp ings. (2025). Main AI T ends in
Educa ion (2025). Re ie ed om
h ps://sp ingsapps.com/knowledge/ma
in-ai- ends-in-educa ion-2024
14. TeachBe e .ai. (2025). The Fu u e o
Educa ion wi h AI: 2025 - A Repo
on Challenges and Oppo uni ies.
Re ie ed om
h ps:// eachbe e .ai/ he- u u e-o -
educa ion-wi h-ai-2025/
15. UNESCO. (2025). A i icial
in elligence in educa ion: Guidance o
policy-make s. Re ie ed om
h ps://www.unesco.o g/en/digi al-
educa ion/a i icial-in elligence
