In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 61
Teache Pe cep ions o he Use o Gene a i e AI Tools in ESL
Class ooms: A Re iew o S udies
Sankaja Ama awee a1, D . Lishan hi Wijewa dhene2
1(Depa men o Linguis ics, Facul y o Humani ies and Sciences/ S i Lanka Ins i u e o In o ma ion
Technology, S i Lanka)
2(Business Communica ion Uni , Facul y o Managemen S udies and Comme ce/ Uni e si y o S i
Jayewa denepu a, S i Lanka)
Abs ac : The in eg a ion o Gene a i e A i icial In elligence (GenAI) ools in English as a Second Language
(ESL) class ooms has gained inc easing a en ion in ecen yea s. While hese ools o e nume ous ad an ages,
such as pe sonalized lea ning, au oma ed eedback, and enhanced s uden engagemen , hei e ec i eness la gely
depends on eache s' pe cep ions and willingness o in eg a e hem in o hei eaching p ac ices. This li e a u e
e iew examines exis ing esea ch on eache pe cep ions o GenAI in ESL class ooms, ocusing on bo h he
po en ial bene i s and he challenges associa ed wi h i s implemen a ion. The e iew explo es hemes such as
pedagogical applica ions, eache a i udes, echnological ba ie s, and e hical conce ns. Findings sugges ha
while many educa o s ecognize he po en ial o GenAI o imp o ing ins uc ional e iciency and s uden
au onomy, conce ns ela ed o accu acy, digi al li e acy, and he isk o o e - eliance on AI-gene a ed con en
pe sis . Addi ionally, a lack o ins i u ional suppo and eache aining emains a signi ican ba ie o
widesp ead adop ion. The e iew highligh s gaps in cu en esea ch, pa icula ly in non-Wes e n ESL con ex s,
and emphasizes he need o u he s udies on p o essional de elopmen p og ammes and long- e m impac s o
he in eg a ion o GenAI in language lea ning.
Keywo ds: English as a Second Language (ESL), Gene a i e A i icial In elligence (GenAI), Cha GPT,
Technology Enhanced Language Lea ning (TELL), Mobile Assis ed Language Lea ning (MALL), Compu e
Assis ed Language Lea ning (CALL)
1. INTRODUCTION
Changes o he wo ld, ela ed o any ield, has a capaci y in a ec ing he ield o Educa ion. Thus, a eas such as
Technology Enhanced Language Lea ning (TELL), Mobile Assis ed Language Lea ning (MALL) and Compu e
Assis ed Language Lea ning (CALL) we e de eloped. Added o he e ol ing en i onmen o educa ional
echnology, Gene a i e A i icial In elligence (GenAI) has e ol ed, as a ans o ma i e ool in he language
eaching and lea ning con ex s.
Highly de eloped algo i hms and models as Cha GPT, Mic oso Copilo , Gemini and ela ed models
which a e capable o gene a ing human-like ex s, images and o he mul imedia con en o igina e om gi en
p omp s a e gene ally e e ed as GenAI ools. One o he ea lies de ini ions o AI gi en by G ewal, as ci ed in
[1], is “ he science and enginee ing o making in elligen machines”, implying ha in elligence can be modelled
by a machine. This de ini ion go b oadened g adually in e p e ing AI as in elligen agen s ha lea n om he
en i onmen and pe o m ac ions [2]. By 2023, i has e ol ed e lec ing ad ancemen s in echnology, changes in
In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 62
unde s anding and aiming shi s on ocus. Ox o d English Dic iona y [3] de ines i as " he heo y and de elopmen
o compu e sys ems able o pe o m asks no mally equi ing human in elligence, such as isual pe cep ion,
speech ecogni ion, decision-making, and language ansla ion."
English as a Second Language (ESL) eaching and lea ning con ex s can u ilise hese ools as p ac ical
applica ions ha assis in p epa ing pe sonalised lea ning ma e ials, au oma ing lesson plans and ools ha p o ide
ins an ly ailo ed eedback on he use o language and o many o he language lea ning/ eaching needs. As an
example, ce ain eaching p omp s in an ESL speaking class oom such as au hen ic, pe sonalised con e sa ions,
ela ed o di e en scena ios can be con enien ly gene a ed using an AI ool as Cha GPT, which sa es he whole
ime spen in b owsing o sea ching se e al books o o he ma e ials. Such inno a i e uses o GenAI ools e lec
a b oade shi in mode n language eaching, whe e he emphasis is inc easingly placed on lea ne -cen e ed,
echnology- enhanced ins uc ion.
One o he c i ical ac o s ha equi es a en ion in in eg a ing echnology in Educa ion is he pe cep ions
o he educa o s. Thei iews on GenAI a e mul i ace ed, including bo h posi i e and nega i e iewpoin s. On one
dimension, many eache s ecognise he po en ial bene i s o u ilising AI ools in enhancing lea ne educa ion on
he o he hand, conce ns pe sis in accu acy, and eliabili y o he gene a ed con en , po en ial o e - eliance on
echnology, e hical challenges ela ed o da a p i acy and he need o p ope aining on he e ec i e use o AI
ools. These mixed, ideas highligh he di icul y in implemen ing new echnologies in educa ional se ings.
Conduc ing a comp ehensi e li e a u e e iew helps o map he cu en s a e o esea ch on GenAI in
ESL con ex s. I enables us o iden i y he b ead h and dep h o s udies ha ha e examined he bene i s and
challenges o AI in eg a ion, he eby e ealing ecu ing hemes, me hodological app oaches, and no able gaps.
By syn hesizing pas indings, his e iew no only alida es he signi icance o explo ing eache pe cep ions bu
also p o ides a amewo k o unde s anding how educa o s ha e esponded o hese echnological inno a ions
o e ime. This c i ical e alua ion is highly impo an in in o ming u u e esea ch di ec ions and o guiding
policy make s and p ac i ione s in de eloping e ec i e s a egies o AI in eg a ion in language eaching.
1.1 Resea ch Objec i es
The p ima y objec i e o his li e a u e e iew is o explo e how ESL eache s pe cei e he use o Gene a i e AI
ools in hei class ooms. Speci ically, his e iew aims o:
1) Examine he posi i e pe cep ions, such as imp o ed ins uc ional e iciency, pe sonalised lea ning
expe iences, and enhanced s uden engagemen .
2) In es iga e he conce ns and challenges oiced by eache s, including issues o accu acy, e hical
implica ions, and he need o adequa e aining.
3) Iden i y gaps in cu en esea ch, pa icula ly in di e se educa ional se ings, o ecommend a eas o
u he s udy.
1.2 Theo e ical F amewo k
This s udy is g ounded in a mul i ace ed heo e ical amewo k ha in eg a es he Technology Accep ance Model
(TAM), Cons uc i is Lea ning Theo y, and Mobile Assis ed Language Lea ning (MALL) o cla i y eache
pe cep ions and pedagogical impac o Gene a i e AI (GenAI) ools in ESL class ooms.
1.3 Technology Accep ance Model (TAM)
Technology Accep ance Model (TAM) se es as a ounda ional amewo k o unde s anding how eache s decide
o adop and in eg a e echnology in o hei eaching p ac ices. Da is desc ibes ha he accep ance o new
echnology is p edominan ly d i en by wo ac o s – “pe cei ed use ulness” and pe cei ed ease o use” [4]. In
his con ex o ESL eaching, i he eache s a e awa e o he bene i s o using echnology, o example, upda e
lesson planning, ob ain eliable language eedback, and inc ease s uden engagemen , he eache s will be inclined
In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 63
o use echnology in hei class ooms. This model also emphasizes ha any issues ound in he echnical ools can
ail in accep ance by he eache s. This s udy in ends o explo e how he ac o s can in luence eache pe cep ions
o GenAI ools, whe he hey will be accep ed as sus ainable and esou ce ul ools o class oom in eg a ion.
1.4 Cons uc i is Lea ning Theo y
As p esen ed by Piage [5] and Vygo sky [6] lea ne s accumula e knowledge h ough hei expe iences in li e and
e lec i e p ac ice o he ac i ely lea ned con en . In eg a ion o GenAI ools in hei ac i e lea ning p ocess can
p o ide hem wi h con ex - ich, in e ac i e au hen ic exe cises. Fo example, in an ESL class oom, he p omp s
which a e used in a speaking ac i i y can be gene a ed wi h GenAI ools shaping and molding hem a he eache ’s
and lea ne ’s disc e ion. This p omo es language acquisi ion consis en wi h cons uc i is p inciples. U ilising
his concep enables lea ne s o os e deepe unde s anding and e ain language skills, making he lea ning p ocess
engaging and e ec i e.
1.5 Mobile Assis ed Language Lea ning
Mobile Assis ed Language Lea ning (MALL) s esses he impo ance o accessible, on-demand lea ning allowing
lea ne s o access ma e ials whene e and whe e e possible. U ilising GenAI ools in he mobile de ices in ESL
lea ning inc eases he bene i s o MALL by p o iding hem wi h eal-li e, au hen ic, pe sonalized language
lea ning exe cises and p ac ice ou side he adi ional language lea ning class ooms. This con inuous lea ning
model aligns wi h mode n educa ional ends owa ds lexibili y and au onomy, u he enhancing he o e all
lea ning expe ience. This esea ch examines how mobile-based GenAI applica ions can enhance lea ning beyond
he class oom and suppo a mo e in eg a ed, li elong language acquisi ion p ocess.
As a whole, hese h ee amewo ks p o ide a comp ehensi e se ing in examining eache pe cep ions
o GenAI in ESL se ings. The ac o s ha in luence echnology adop ion in ESL se ings a e examined wi h TAM,
while he possibili y o c ea ing an engaging and au hen ic expe ien ial lea ning en i onmen is highligh ed by
Cons uc i is Lea ning Theo y [5]. MALL is u ilized he e in examining he impo ance o mobile echnology in
making hese inno a ions accessible and e ec i e o con inuous language lea ning. These in eg a ed heo e ical
amewo ks highligh he mul i-dimensional impac o GenAI applica ions on ESL eaching and aids in p o iding
s a egies in e ec i e implemen a ion o hem in ESL lea ning and eaching en i onmen s.
1.5 Me hodology o he Li e a u e Re iew
In o de o examine eache pe cep ions on he use o GenAI ools, a sys ema ic app oach was adop ed in selec ing
and analysing ele an li e a u e. This e iew u ilises pee - e iewed jou nal a icles, books and con e ence
p oceedings published in 2023 and 2024 o ep esen he ecen shi s and de elopmen s in he ield.
The selec ion o a icles was based on keywo d sea ches in h ee da abases: Google Schola , Scopus and
ELSEVIER. The e ms used o he sea ches we e – Gene a i e AI in ESL Teaching, Teache Pe cep ions o AI
in Educa ion, Teache Pe cep ions o Inco po a ing AI in English Language Teaching. The s udies we e
conside ed based on hei ele ance o he esea ch objec i es ocusing on hei empi ical indings, heo e ical
applica ions and pedagogical implica ions.
The hema ic analysis was based on he majo hemes such as “ eache a i udes owa ds GenAI, bene i s
and challenges o he inco po a ion o GenAI and he ole played by AI ools in AI-assis ed language lea ning.
This s uc u ed amewo k can assis in unde s anding he exis ing body o knowledge and iden i ying gaps o
u u e esea ch.
II. Re iew o Li e a u e
This e iew syn hesizes key hemes, including he ole o GenAI in ESL eaching, eache pe cep ions,
pedagogical bene i s and challenges, and empi ical s udies conduc ed in ESL se ings u ilising GenAI ools. The
In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 64
analysis o he exis ing esea ch sec ion highligh s ends, heo e ical pe spec i es, and gaps, p o iding insigh s
in o he e ol ing ole o AI in ESL educa ion.
2.1 The Role o Gene a i e AI in ESL Teaching
Since he a i al o he GenAI ools, he way ha hey can be inco po a ed in ESL class ooms has been esea ched.
Se e al non-empi ical s udies [7] – [9] ha e esea ched he ole o GenAI ools in ESL eaching.
C eely [7] highligh s se e al key oppo uni ies ha can be ob ained om GenAI ools in language
lea ning. In his concep ual analysis he p oposed a concep ual model in in eg a ing GenAI ools in language
educa ion o language educa o s and esea che s. Al hough an exac “model” is no de ailed, C eely ad oca es a
balanced and e hical app oach in u ilising AI by syn hesising oppo uni ies and challenges o GenAI while
add essing i s limi a ions and e hical conside a ions. This amewo k suppo s pe sonalized leaning expe iences,
dynamic con en gene a ion and immedia e eedback, aiming a enhancing language acquisi ion and li e acy skills.
In his concep ual explo a ion, I onsi [8] o e s insigh s and ecommenda ions on he use o GenAI in
English language eaching. GenAI enhances English language lea ning by p o iding i id a he same ime
ele an con en , which aids he lea ne s acqui e and use he language mo e e ec i ely. I can also boos s uden
engagemen h ough in e ac i e and pe sonalized expe iences. Fu he mo e, AI has he abili y o adap lessons o
ma ch indi idual lea ning s yles and needs, p o iding be e lea ning ou comes.
Liao e al. [9] use a case analysis using he ins uc ion- eply me hod in o de o in es iga e he po en ial
o in eg a ing GenAI in lea ning and p ac ice o he ou language skills – lis ening, eading, w i ing and speaking.
They use Cha GPT in his s udy, p o iding p omp s and analysed he esponses. They iden i ied nine co e abili ies
o AI which con ibu e o language lea ning as – na u al language p ocessing, adap i e lea ning, speech
ecogni ion and syn hesis, a en ion mechanisms, ze o-sho lea ning, p e- ained models, in elligen
ecommenda ion, ein o cemen lea ning and c oss-lingual lea ning. These ea u es enable AI ools o deli e
pe sonalized eedback, simula e eal communica ion, and o e di e se lea ning esou ces and expe iences.
Conside ing he abo e h ee esea ch on he ole played by GenAI in ESL se ings, i is e iden ha
C eely [7] p o ides a concep ual amewo k in in eg a ing AI in a balanced manne while espec ing e hics. I onsi
[8] highligh s p ac ical bene i s in enhancing lea ne engagemen and imp o ing ou comes h ough pe sonalized,
a ied con en . Bo h he esea che s emphasise he need o ca e ul implemen a ion, hough C eely [7] is mo e
ocused on e hical syn hesis, whe eas I onsi [8] is pa ial owa ds pedagogical applica ion o AI ools. Liao e al.
[9] on he o he hand, o e a mo e empi ical app oach by iden i ying nine co e AI capabili ies. Unlike C eely’s
[7] concep ual emphasis and I onsi’s [8] class oom-le el insigh s, Liao e al. [9] b idge heo y and p ac ice h ough
eal- ime AI ins uc ion. Toge he hese s udies p esen a holis ic iew o using GenAI in language: C eely [7]
ou lines he e hical and concep ual amewo k, I onsi [8] si ua es i in pedagogical p ac ice, and Liao e al. [9]
demons a e i s unc ional e ec i eness, highligh ing he impo ance o aligning AI capabili ies wi h e hical and
lea ne -cen e ed eaching s a egies.
2.2 Teache Pe cep ions o Gene a i e AI
A g owing body o esea ch highligh s he ans o ma i e po en ial o GenAI ools o enhance eaching and
lea ning p ocess, os e lea ne au onomy and p o ide adap i e, pe sonalised suppo o s uden s. S udies by
Chung & Jeong [10], Ta azoli [11], AI-kh esheh [12], Kal a [13], and Nowa an e al. [14] collec i ely ea u e he
posi i e impac o AI powe ed echnologies when hough ully in eg a ed in o ESL class ooms. These s udies
highligh how GenAI can suppo lesson planning, gene a e engaging con en , o e eal- ime eedback and
p omo e language de elopmen beyond adi ional class oom bounda ies. Align wi h hese indings, he ollowing
sec ion explo es he posi i e pe cep ions and pedagogical ad an ages o employing GenAI in ESL educa ion.
In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 65
2.2.1 Posi i e Pe cep ions
Chung & Jeong [10] examined pe cep ions o 134 Chinese p ese ice eache s (PSTs) on inco po a ing AI ools
in English language eaching. This s udy ound ha PSTs we e amilia wi h he selec ed AI ool – Cha GPT and
ecognized hei desi e o ha e p o essional de elopmen in his a ea since many lacked a deep unde s anding o
he unc ions and how hey can be u ilised o achie e maximum bene i o i . The pa icipan s o his s udy
acknowledged ha AI can be used in imp o ing g amma and w i ing skills and no iced i s abili y o suppo
pe sonalised lea ning wi hou educing in e es in adi ional me hods. They u he exp essed a high le el o
accep ance owa d using AI in eaching, ecognising i s impo ance in inc easing e iciency, sa ing ime in lesson
planning and assessmen , and imp o ing s uden s’ comp ehension o lea ning ma e ials.
In o de o explo e he po en ial o u ilising GenAI by English language eache s o o e come challenges
in eaching English, Ta azoli [11] conduc ed a s udy wi h 23 pa icipan s. The s udy was conduc ed as a quali a i e
case s udy examining he e ec i eness o Cha GPT as a easible ool in ESL eaching. The s udy iden i ied se e al
plus poin s in inco po a ing Cha GPT as accessibili y in high quali y ma e ials, inclusi i y o e e y ype o
s uden s, pe sonalised lea ning, inco po a ing enhanced eaching and lea ning me hods and inc easing s uden
engagemen and mo i a ion.
Ano he s udy explo ed global English language eache ’s pe cep ions o in eg a ing AI ools in o ELT
(English Language Teaching) h ough a quali a i e design in ol ing 46 pa icipan s om 39 coun ies [12]. Using
open-ended ques ionnai es, he esea ch iden i ied key pedagogical oppo uni ies such as eal- ime eedback,
pe sonalized lea ning, enhanced s uden engagemen , and suppo o w i ing and c i ical hinking. Teache s also
alued Cha GPT's ole in helping quie e s uden s o pa icipa e mo e con iden ly in he class oom ac i i ies.
Looking o wa d, eache s ad oca ed o a hyb id ELT model whe e AI complemen s adi ional ins uc ion,
highligh ing he need o cu iculum edesign, eache aining, and he de elopmen o s uden s’ digi al, c i ical,
and emo ional li e acies o ha ness he ull po en ial o AI ools like Cha GPT.
Kal a [13] explo ed language eache ’s pe cep ions o in eg a ing AI ools in he class oom h ough a
su ey o 208 educa o s. The indings indica ed ha while mos eache s we e mode a ely amilia wi h AI ools
like G amma ly and Cha GPT, only a ew use hem egula ly. The pa icipan s ha e ecognised he po en ial o
AI ools o enhance engagemen , pe sonalise lea ning, and p o ide immedia e eedback. The majo i y o he
pa icipan s ha e shown a posi i e a i ude and willingness o adop AI ools, p o ided ha adequa e aining and
ins i u ional suppo we e in place.
Nowa an e al. [10] in hei s udy explo e how Technology Enhanced Language Lea ning (TELL) is
used in language eaching, pe spec i es o he educa o s on u ilising AI, implica ion policies, and challenges o i .
Se en English language eache s who a e using AI ools a e included as he pa icipan s o his quali a i e s udy.
The indings o his s udy indica ed mo e posi i e a i udes owa ds inco po a ing AI ools in he class oom due
o he bene i s as enhancing ma e ial de elopmen , main aining an engaged and en husias ic class oom and
con inuing e ec i e assessmen s. Fu he , “a hyb id model” – balancing AI in eg a ion wi h human in e ac ion
was encou aged.
Thus, he abo e s udies, highligh nume ous posi i e e ec s o in eg a ing AI ools such as Cha GPT
in o English language eaching. A signi ican bene i iden i ied is he imp o emen o s uden s’ g amma and
w i ing skills h ough eal- ime eedback and language modelling ea u es [10], [12]. Ano he majo ad an age o
AI echnologies is he acili a ion in pe sonalised lea ning by adap ing o indi idual s uden needs, enabling
di e en ia ed ins uc ion while main aining engagemen wi h adi ional me hods [10] – [13]. The educa o s in
hese s udies also epo ed AI ools aid in inc eased s uden mo i a ion and pa icipa ion wi h AI, encou aging
e en quie e lea ne s o engage mo e ac i ely in class oom ac i i ies [10] – [14]. Mo eo e , he s udies [10] and
[14] in o m ha in eg a ion o AI ools can enhance eaching e iciency by sa ing ime in lesson planning and
c ea ing e ec i e assessmen s. Also, GenAI ools can p o ide access o high quali y lea ning ma e ials [11], [14],
In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 66
which a e inclusi e o di e se lea ne p o iles [11]. This s udy u he claims ha AI ools aid in p o iding
enhanced eaching me hods.
I is also indica ed in he abo e s udies ha while a conside able numbe o eache s a e mode a ely
amilia wi h he AI ools, many in o med o posi i e a i udes and eage ness o adop hem, especially when
adequa e aining and ins i u ional suppo a e gi en [13], [14]. Two s udies also p omo e hyb id eaching models
whe e AI complimen s adi ional ins uc ion, emphasizing he need o cu iculum edesign and eache aining
o ully ha ness he pedagogical po en ial o AI ools [12], [14]. Las ly, he impo ance o con inuous p o essional
de elopmen was highligh ed [10], as many educa o s seek o deepen hei unde s anding and e ec i e use o AI
in language ins uc ion.
2.2.2 Nega i e Pe cep ions
In all he abo e s udies, in line wi h he mul iple bene i s o AI in eg a ion, conside able conce ns we e also
discussed. Al-kh esheh [12] sugges s se e al conce ns o using cha GPT in ESL class ooms. One main challenge
is i s inabili y o cap u e some essen ial aspec s o language lea ning - cul u al con ex s, emo ional dep h and
con ex speci ic meanings. Due o i s ex - based na u e, Cha GPT lacks suppo o o al skills such as
p onuncia ion and lis ening. Al-kh esheh [12] also no ed he ailu e o Cha GPT o simula e spon aneous, eal-
li e communica ion which is impo an in imp o ing au hen ic language use. He u he examines he iew o
many educa o s - o e eliance on Cha GPT and o he AI ools, which po en ially diminish he s uden s’ c i ical
hinking and p oblem-sol ing abili ies.
Addi ionally, he use o AI-gene a ed esponses may supp ess s uden s’ c ea i i y, especially in asks
ha equi e o iginal exp ession. The digi al di ide poses ano he challenge, as no all s uden s and eache s ha e
equal access o echnology, o he digi al skills needed o use i e ec i ely. Finally, he in eg a ion o Cha GPT
o en clashes wi h exis ing cu icula and e eals a gap in eache aining, unde sco ing he need o p o essional
de elopmen and pedagogical alignmen .
Kal a [13] wi h he numbe o op imis ic iews on he use o AI usage in ESL class ooms, obse es
se e al conce ns among ins uc o s ega ding he in eg a ion o AI ools in ESL class ooms. O e eliance is one
o he p ima y issues ha was obse ed in he s udy which hinde s he lea ne s’ abili y o hink c i ically and
ce ain cogni i e unc ions as b ains o ming. This encou ages plagia ism. Kla a’s [13] esea ch pa icipan s also
doub ed he eliabili y o AI gene a ed in o ma ion, logical a gumen s and ci ing. Skep icism on e hical conce ns
pa icula ly ela ed o da a p i acy and academic in eg i y was also highligh ed. Fu he mo e, eluc ance in
adop ing new echnologies, un amilia i y wi h AI ools and lacking p ope aining we e iewed as nega i e
pe cep ions. Finally, hei doub s abou eache oles in an AI-assis ed lea ning en i onmen we e aised.
Ce ain app ehensions o AI usage despi e i s’ a ou able iews we e gi en by Nowa an e al. [14] in
hei s udy. P edominan ly hey emphasise he lack o c i ical e lec ion, p o essional de elopmen , policy
guidance in u ilising AI applica ions. They also highligh ed depe sonaliza ion ( ha is limi ed human connec ion)
in AI solu ions, lack o a obus policy amewo k o e hics and da a p i acy.
Simila ly, he s udy [10] e eals se e al conce ns among Chinese P e-se ice eache s ega ding he
in eg a ion o AI applica ions. P ima ily hei conce n was on he o e dependency on au oma ed suppo which
obs uc lea ne ’s c ea i i y and c i ical hinking abili ies. The e we e also ques ions on he au hen ici y o AI-
acili a ed in e ac ions, obse ing ha GenAI’s s uggles in eplica ing cul u al and con ex ual nuances essen ial
o meaning ul language lea ning. Fu he , he esea che s highligh ed e hical conce ns ela ed o da a p i acy,
issues in academic in eg i y, and he p oduc ion o biased o inaccu a e in o ma ion as some o he majo conce ns
o he in eg a ion o AI ools in ESL se ings. This s udy u he poin s ou he ine iciency o AI applica ions in
suppo ing language skills as speaking and lis ening. Finally, he un amilia i y and lack o unde s anding on he
e ec i e use o AI applica ions by he language eache s we e emphasised.
In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 67
Liao e al. [9] also in hei s udy highligh s limi a ions o he in eg a ion o AI as he p oduc ion o
mechanical and unna u al language, high use p o iciency equi emen s, inconsis en language ecogni ion, and
he p obable language inaccu acies. Thus, hey sugges he need o cau ious implemen a ion o AI ools in ESL
con ex s.
The e o e, i is clea ha a ange o c i ical conce ns we e add essed in he abo e i e s udies in he
in eg a ion o AI ools in ESL eaching and lea ning en i onmen s. One o he common issues discussed in he
s udies is he o e - eliance on AI ools by he lea ne s which esul s in de e io a ing he lea ne ’s c ea i i y, c i ical
hinking and p oblem-sol ing abili ies [12], [13],[10], while encou aging plagia ism [13]. Ano he majo conce n
which is highligh ed is he ailu e o AI ools (pa icula ly ools as Cha GPT) in suppo ing essen ial language
skills such as speaking, lis ening and spon aneous, eal-li e communica ion skills which a e essen ial in au hen ic
language skills [12], [10]. The limi a ions in eplica ing cul u al nuances, emo ional dep h and indica ing con ex -
speci ic meanings con inue o limi he e icacy o AI ools in c ea ing meaning ul in e ac ions in language
lea ning [12], [10]. Reliabili y o he con en including da a p i acy issues, academic in eg i y and e hical conce ns
ha e been ex ensi ely discussed by Kal a [13], Chung & Jeong [10] and Nowa an e al. [14]. Also, lack o
amilia i y wi h AI ools, insu icien eache aining, and hesi ancy in adop ing new echnologies a e conside ed
as no able ba ie s in he in eg a ion o AI applica ions in ESL se ings [12], [10], [13]. These challenges a e
in ensi ied due o digi al di ide as he e a e dispa i ies in equal access o AI echnologies among he ESL educa o s
and lea ne s [12]. Mechanical and unna u al language ou pu , language inaccu acies and inconsis en ecogni ion
a e also highligh ed as limi a ions o he in eg a ion o AI ools. Thus, ca e ul implemen a ion is sugges ed [9].
Las ly, he lack o powe ul policy guidance and e hical amewo ks u he add o complica e he inco po a ion
o AI in ESL con ex [14]. The e o e, he abo e conce ns highligh he necessi y o p o essional de elopmen ,
e hical sa egua ds and pedagogical alignmen in con i ming AI echnologies collabo a e a he han impede ESL
eaching and lea ning.
III. Discussion
This pape analysed he eme ging ole o Gene a i e AI (GenAI) applica ions in ESL (English as a Second
Language) educa ion, ocusing on pedagogical applica ions and eache pe cep ions. Ou esea ch ou lines a
pic u e o he bene i s and challenges expe ienced by ESL educa o s in hei eaching and lea ning p ocess du ing
he pas yea . The li e a u e e ealed ha wi h he many ewa ds o he inco po a ion o GenAI ools in enhancing
ESL ins uc ion as pe sonalised lea ning, ob aining eal- ime eedback, high e iciency in lesson planning and
de elopmen o adap i e eaching/lea ning ma e ials, suppo o di e en ia ed ins uc ion which inc eases
s uden engagemen and en husiasm, he eby os e ing lea ne au onomy.
Howe e , he e iew also ma ks c i ical conce ns o he in eg a ion o GenAI in ESL en i onmen s. The
educa o s a e oubled by he po en ial o e - eliance on AI ools which may weaken he lea ne s’ essen ial
cogni i e skills as c ea i i y, c i ical hinking and p oblem-sol ing abili ies. The e iew also highligh s he
supp ession o o iginal exp ession and inc eased endency o plagia ism pa icula ly ela ed o w i ing ac i i ies.
Mo eo e , he ine ec i eness o GenAI o e ec i ely suppo speaking and lis ening skills, along wi h i s
limi a ions in cap u ing cul u al and con ex ual nuances, dec eases i s e ec i eness in suppo ing holis ic language
compe ence. Conce ns ela ed o he accu acy and bias o he AI-p oduced con en , da a p i acy, academic
in eg i y and he digi al di ide u he adds o he issues o GenAI ools which esul in u he delaying he
in eg a ion o i in he ESL class ooms.
Ano he ecu ing heme is he eadiness o he eache s o adop AI ools. While many educa o s
demons a e a willingness o use GenAI, he lack o adequa e aining and ins i u ional suppo is a majo ba ie .
Teache s equi e ongoing p o essional de elopmen o na iga e e hical conside a ions, maximize pedagogical
bene i s, and align AI in eg a ion wi h cu iculum goals.
In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 68
In addi ion, conce ns abou he ede ini ion o eache oles in AI-assis ed en i onmen s indica e a need
o clea e amewo ks ha posi ion GenAI as a suppo i e, a he han a eplacemen ool in educa ion.
The e iewed s udies also emphasise he impo ance o a hyb id eaching model, whe e AI is used o
complemen adi ional me hods. This app oach helps p ese e he human elemen in language eaching which is
essen ial o add essing emo ional, cul u al, and in e pe sonal dimensions o lea ning while le e aging he
s eng hs o GenAI o enhance ins uc ional e ec i eness.
IV. Limi a ions
This esea ch shows ha eache s mainly used AI o lesson planning and ma e ials de elopmen , wi h minimal
use in assessmen and eal- ime class oom ac i i ies. This na ow ocus may ully cap u e he b oade po en ial o
AI in language lea ning. Secondly, his s udy is based on cu en pe cep ions and p ac ices. I lacks long- e m da a
showing how AI in eg a ion impac s language p o iciency o e ime. The pa icipan s o his s udy possess limi ed
expe ience in he ull po en ial o he AI ools; hus, he e ec i eness o AI is cons ained by he use ’s lack o
expe ise o unde s anding. Addi ionally, he s udies conside ed o his esea ch a e mainly o Asian, A abian
and Aus onesian. The e o e, he esul s may no be gene alizable o o he con ex s wi h di e en echnological
in as uc u e, pedagogical cul u es o policy en i onmen s. Al hough his e iew acknowledges conce ns such as
depe sonalisa ion and he need o e hical guidelines, i does no deeply explo e hese issues o p opose conc e e
solu ions. The ocus o his s udy is mainly on eache s’ pe cep ions. A lack o di ec inpu om s uden s o e looks
c i ical insigh s in o how lea ne s expe ience and pe cei e AI in language educa ion.
Fu he mo e, his s udy ocusses mainly on he AI applica ion – Cha GPT. Howe e , conside ing he
apid pace o AI de elopmen , he e a e nume ous o he AI applica ions which can bene i ESL educa ion. Finally,
he numbe o a icles selec ed o his s udy is limi ed o he pas couple o yea s and depending on he
pe spec i es o he educa o s only. This can esul in lack o iangula ion.
V. Sugges ions
Gi en he e ol ing na u e o GenAI in ESL educa ion, se e al a enues o u u e esea ch eme ge. Mo e empi ical
and longi udinal s udies a e needed o e alua e he sus ained impac o GenAI ools on a ious aspec s o language
acquisi ion, including speaking and lis ening skills. In o de o ill he gap in he skills, in es iga ing e ec i e
aining amewo ks ha empowe eache s o use GenAI e hically and pedagogically is iden i ied as a necessi y.
Fu u e s udies also need o ocus on de eloping obus e hical guidelines and ins i u ional policies o add ess
conce ns ela ed o da a p i acy, academic in eg i y, and equi able access. Mo eo e , i is impo an o explo e
how GenAI ools can be adap ed o be e align wi h lea ne di e ences, including p o iciency le els, cul u al
backg ounds and lea ning p e e ences. Mos impo an ly, e alua ions o he e ec i eness o blended eaching
models which in eg a es human in e ac ion and AI assis ance can o e insigh s in o sus ainable and scalable ESL
p ac ices.
VI. Conclusion
The in eg a ion o A i icial In elligence (AI) in language educa ion p esen s signi ican oppo uni ies o enhance
cu iculum design, pedagogy, and assessmen p ac ices. This s udy highligh s ha while AI is p edominan ly used
by eache s o lesson planning and ma e ials de elopmen , i s applica ion in class oom ins uc ion and assessmen
emains limi ed. Teache s pe cei e AI as a aluable ool o p omo ing pe sonalised lea ning, imp o ing
ins uc ional e iciency, and suppo ing s uden au onomy. Howe e , he e ec i eness o AI in language lea ning
cu en ly is hinde ed by gaps in eache and s uden eadiness, including limi ed c i ical e lec ion, lack o
p o essional de elopmen and insu icien aining in AI ools.
The indings also unde sco e he impo ance o main aining human in e ac ion in he language lea ning
p ocess. Al hough AI can o e adap i e eedback and pe sonaliza ion, he absence o emo ional engagemen and
In e na ional Jou nal o A s and Social Science www.ijassjou nal.com
ISSN: 2581-7922,
Volume 8 Issue 11, No embe 2025
Sankaja Ama awee a Page 69
au hen ic communica ion may lead o depe sonalized lea ning expe iences. The e o e, a balanced app oach ha
combines he s eng hs o AI wi h he i eplaceable human elemen is essen ial.
To ealize he ull po en ial o AI in language educa ion, a comp ehensi e and e hical amewo k is
necessa y. This includes aligning AI in eg a ion wi h cu icula goals, de eloping eache compe encies,
e hinking assessmen me hods, and es ablishing clea policies on esponsible AI use. Ongoing collabo a ion
among educa o s, policymake s, and echnology de elope s will be i al o ensu e AI is implemen ed in ways ha
uphold he co e alues o language lea ning. As echnology con inues o e ol e, language educa o s mus ake he
lead in le e aging AI o c ea e inclusi e, engaging, and e ec i e lea ning en i onmen s.
Re e ences
[1] P. D. S. G ewal, “A c i ical concep ual analysis o de ini ions o a i icial in elligence as applicable o
compu e enginee ing,” IOSR J. Compu . Eng., ol. 16, no. 2, pp. 9–13, 2014. [Online]. A ailable:
h ps://doi.o g/10.9790/0661-16210913
[2] S. Russell and P. No ig, A i icial In elligence: A Mode n App oach, 3 d ed. Uppe Saddle Ri e , NJ:
P en ice Hall, 2010.
[3] Ox o d English Dic iona y, “A i icial in elligence,” Ox o d English Dic iona y, 2023. [Online]. A ailable:
h ps://www.oed.com/
[4] F. D. Da is, “Pe cei ed use ulness, pe cei ed ease o use, and use accep ance o in o ma ion echnology,”
MIS Q., ol. 13, no. 3, pp. 319–340, 1989. [Online]. A ailable: h ps://doi.o g/10.2307/249008
[5] J. Piage , The Psychology o In elligence, M. Pie cy and D. E. Be lyne (T ans.), (London, U.K.: Rou ledge
& Kegan Paul, 1966).
[6] L. S. Vygo sky, Mind in Socie y: The De elopmen o Highe Psychological P ocesses, M. Cole, V. John-
S eine , S. Sc ibne , and E. Soube man (Eds. & T ans.), (Camb idge, MA: Ha a d Uni . P ess, 1978).
[7] E. C eely, “Explo ing he ole o gene a i e AI in enhancing language lea ning: Oppo uni ies and
challenges,” In . J. Changes Educ., ol. 1, no. 3, 2024. [Online]. A ailable:
h ps://doi.o g/10.47852/bon iewIJCE42022495
[8] C. S. I onsi, “Explo ing he po en ial o gene a i e AI in English language eaching,” in Facili a ing Global
Collabo a ion and Knowledge Sha ing in Highe Educa ion wi h Gene a i e AI, pp. 162–185, 2024.
[Online]. A ailable: h ps://doi.o g/10.4018/979-8-3693-0487-7.ch007
[9] H. Liao, H. Xiao, and B. Hu, “Re olu ionizing ESL eaching wi h gene a i e a i icial in elligence—Take
Cha GPT as an example,” In . J. New De elop. Educ., ol. 5, no. 20, 2023. [Online]. A ailable:
h ps://doi.o g/10.25236/ijnde.2023.052008
[10] J. Y. Chung and S.-H. Jeong, “Explo ing he pe cep ions o Chinese p e-se ice eache s on he in eg a ion
o gene a i e AI in English language eaching: Bene i s, challenges, and educa ional implica ions,” Online
J. Commun. Media Technol., ol. 14, no. 4, e202457, 2024. [Online]. A ailable:
h ps://doi.o g/10.30935/ojcm /15266
[11] D. Ta azoli, “Explo ing he po en ial o gene a i e AI in democ a izing English language educa ion,”
Compu . Educ.: A i . In ell., ol. 7, 100275, 2024. [Online]. A ailable:
h ps://doi.o g/10.1016/j.caeai.2024.100275
[12] M. H. Al-kh esheh, “B idging echnology and pedagogy om a global lens: Teache s’ pe spec i es on
in eg a ing Cha GPT in English language eaching,” Compu . Educ.: A i . In ell., ol. 6, 100218, 2024.
[Online]. A ailable: h ps://doi.o g/10.1016/j.caeai.2024.100218
[13] R. Kal a, “Explo ing eache s’ pe cep ions owa d he in eg a ion o AI ools in he language class oom,”
NIDA J. Lang. Commun., ol. 29, no. 45, pp. 21–32, 2024.
[14] A. No awan, S. A. Walke , and O. Ikeda, “The new ace o echnology-enhanced language lea ning (TELL)
wi h a i icial in elligence (AI): Teache pe spec i es, p ac ices, and challenges,” J. Engl. Acad. P o .
Commun., ol. 10, no. 1, pp. 1–18, 2024. [Online]. A ailable: h ps://doi.o g/10.25047/jeapco. 10i1.4565
