In oduc ion
JC Pene
Newcas le Uni e si y, Uni ed Kingdom
1 Why his book?
Since OpenAI launched Cha GPT back in No embe 2022, gene a i e a i icial
in elligence (GenAI) has ac ed as a majo dis up o in he ansla ion indus y
and beyond. O cou se, in any gi en indus y dis up o s a e no in insically
good o bad. They me ely ac as game change s, good and bad. Also commonly
associa ed wi h “cha bo s”, GenAI ools ce ainly a e a game change o he
ansla ion indus y and, consequen ly, o ansla o educa ion oo. Using deep
lea ning o ain on as co po a, no only do LLMs ha e he abili y o “gene a e
ex which is o en indis inguishable om ex w i en by a human” (Moo kens
e al. 2025: 188), bu hey can do much mo e han jus ansla e ex s. In his
book, we shall he e o e make a dis inc ion be ween a i icial in elligence (AI)
and GenAI. In ou con ex , AI e e s b oadly o all AI echnologies, including
neu al machine ansla ion (NMT). GenAI, howe e , e e s o machine lea ning
ools ha gene a e media, including cha -based La ge Language Models (LLMs)
like GPT-4. Since his book deals mos ly wi h ex , he e ms GenAI and LLMs
end o appea synonymously.
Unlike eely a ailable machine ansla ion (MT) ools like Google T ansla e o
DeepL, LLMs can also be p omp ed o adap hei ansla ions o ake he sou ce
ex and/o he a ge ex con ex in o conside a ion – o ee (ini ially). In a
ma e o seconds. I is easy o see how GenAI can po en ially con ibu e o he
u he democ a isa ion o ansla ion by making i eely accessible in con ex s
whe e ansla ion would ha e been nei he p ac ical no a o dable un il now.
This dis up o also p o ides ansla o s wi h new esou ces ha can po en ially
help hem no jus o inc ease he speed and quali y o he ansla ion wo k hey
deli e o clien s, bu also wi h he way hey manage hei ansla ion p ojec s o
JC Pene . 2026. In oduc ion. In JC Pene , Joss Moo kens & Masa u Ya-
mada (eds.), Teaching ansla ion in he age o gene a i e AI: New pa adigm,
new lea ning?, iii–ix. Be lin: Language Science P ess. DOI: 10.5281/zenodo.
17641062
JC Pene
go abou inding new clien s. In he 2024 edi ion o he annual ELIS (Eu opean
Language Indus y Su ey) epo , o ins ance, he independen ansla o s who
saw GenAI as a posi i e end explained ha hey le e aged i s possibili ies “as a
ool (e.g. e minology ex ac ion), as a sou ce o edi ing wo k, and as a mo i a o
o clien s o choose human ansla ion due o bad AI expe iences” (ELIS 2024:
24).
O cou se, he use o AI cha bo s o ansla ion wo k also comes wi h some
impo an limi a ions ha need ou a en ion. One o hem is ha hey need as
amoun s o da a o lea n, meaning ha hei abili y o ansla e depends la gely
on he amoun o language da a a ailable o ain on o any gi en language.
E en hough LLMs di e om all o he machine lea ning-based MT sys ems in
ha hey can be ained on monolingual da a ins ead o bilingual co po a, hus
inc easing he amoun o a ailable da a hey can ain on, hey emain mos ly
ained on English da a. Because o his, hei ansla ion pe o mance is likely
o be much less con incing o so-called low- esou ce languages compa ed wi h
high- esou ce languages . In addi ion, AI cha bo s a e ained o iden i y and
eplica e language pa e ns. A na u al co olla y o his is ha hey lack “any o m
o genuine comp ehension o he cogni i e p ocesses ha humans use o unde -
s and language and con ex ” (Moo kens e al. 2025: 189). Finally, when GenAI
lacks he da a i needs o p oduce pa o a ansla ion, i can “hallucina e”, i.e.
make up con en in he a ge language.
These limi a ions no wi hs anding, GenAI has s a ed eshaping wha i means
o wo k as a p o essional ansla o in an indus y ha was al eady la gely
echnology-d i en, and whe e a guably au oma ion has long been used as a
way o inc ease he speed and each o ansla ion while educing i s cos . This
has some imes been o he de imen o wo king condi ions and job sa is ac ion
o human ansla o s (see, o ins ance, Lambe & Walke 2024). This, in ac ,
makes i e en mo e u gen o all o us ansla o educa o s o engage wi h GenAI
so ha we can, in he p ocess, e-examine he ole and agency o human ans-
la o s wi hin he ansla ion p ocess h ough ha lens. In o he wo ds, GenAI
should encou age us o e hink c i ically he alue and alues ha humans b ing
o he ansla ion p ocess. As pa o his, we mus ( e-)in e oga e wha we do on
ou ansla ion p og ammes so ha we empowe he nex gene a ion o p o es-
sional ansla o s o achie e he kind o “human-cen ed augmen ed ansla ion”
ha will bene i bo h indi iduals and socie y (O’B ien 2024: 391).
Admi edly, howe e , o some o us ansla o educa o s he p ospec o en-
gaging wi h GenAI may ini ially ha e el somewha o e whelming. Fi s o all,
he iming may no ha e been ideal. OpenAI’s Cha GPT was eleased sho ly a -
e we came ou blinking om a se ies o pandemic- ela ed lockdowns, du ing
i
In oduc ion
which eaching happened mos ly online. Fo some o us, his mean ha ing o in-
e ac wi h (new) echnologies ha we may ha e el we e no always ailo ed o
ou own needs and/o ha we we en’ always com o able using. As a esul ,
mos o us go o expe ience i s -hand he “ echnos ess” many p o essional
ansla o s expe ience when hey a e asked o use ansla ion ools and ech-
nologies hey a e no en i ely com o able wi h (Pene 2024). This is some hing
we should ake se iously, as echnos ess can “ educ[e] pe o mance and ha [m]
indi idual wellbeing” (Koskinen 2020: 146). Wi h i s pedagogical app oach, his
edi ed olume he e o e comes as an a emp o help alle ia e eelings o s ess
among some o us as ansla o aine s.
Ideally, we should all eel empowe ed o engage wi h GenAI on ou p o-
g ammes, whe e i also ac s as a dis up o . Again, his was e lec ed in he
indings o he la es ELIS epo s. In i s 2024 su ey, uni e si y s a anked
GenAI implemen a ion as he mos widely sha ed challenge, wi h close o 90%
o hem seeing i as an issue. This p omp ed he epo ’s au ho s o commen
ha : “Gene a i e AI and how o implemen i in he uni e si ies’ p og ammes
has aken he challenge cha o uni e si y s a by s o m, opping e en hei
conce ns abou he isibili y o he p o ession and he e e nal lack o ime” (ELIS
2024: 27). I , he ollowing yea , isibili y o he p o ession had eclaimed op
spo in he lis o challenges, i was s ill closely ollowed by GenAI implemen-
a ion, which emained a conce n o o e 80% o uni e si y s a (ELIS 2025:
25). Ye , adap and implemen we mus ! This is because, s ill acco ding o he
epo , in 2025 “[a]c ual MT use by language companies has inc eased […] and
eaches now he magic ma k o 50% o handled p ojec s. AI makes i s en y wi h
an imp essi e 34%” (ELIS 2025: 35). In a simila ein, Sla o , an online po al o
language indus y news and esea ch, ecen ly made he ollowing poin :
‘Ska e o whe e he puck is going, no o whe e i is’ has always been good
ad ice o ice hockey. Bu i is ema kably p escien o he language se -
ices indus y as we p epa e o he sunami o dis up ion b ough abou by
gene a i e a i icial in elligence (AI) and la ge language models, o LLMs.
The speed o de elopmen has been ene ic since he elease o OpenAI’s
Cha GPT in No embe 2022. […] I would be a b a e o ecas e o p edic
how he co po a e landscape will look i e yea s om now—bu ope a ing
om whe e he puck is oday isks asking he w ong ques ions and missing
he boa (Welocalize 2023).
E en hough his message was clea ly in ended o language se ice p o ide s,
i ce ainly holds some ele ance o us educa o s. Wha e e we may hink o eel
JC Pene
abou GenAI, we owe i o bo h ou s uden s and o he ansla ion indus y o
engage wi h i , and o do so in a way ha will help hem an icipa e whe e he
puck is going. Howe e , we mus do so in a way ha is e hical and sus ainable as,
a guably, GenAI is much mo e han jus ano he ool in he ansla o ’s oolki .
This is wha his book in ends o help you wi h.
We ha e called his edi ed olume Teaching T ansla ion in he Age o Gen-
e a i e AI: New Pa adigm, New Lea ning. Gi en he e y na u e o GenAI, he
wo d pa adigm, wi h i s his o ical meaning o “pa e n”, seems a he app op i-
a e. Beyond his, he wo d cap u es GenAI’s po en ial o be a game change no
jus in he ansla ion indus y, bu also in ansla o educa ion. Wi h his new
pa adigm, which is cha ac e ised by he eme gence o new ansla ion p ac ices,
comes new lea ning o bo h s uden s and educa o s. This is wha we ha e a -
emp ed o cap u e he e wi h he help o expe s om ac oss he globe. In ac ,
i his edi ed olume was a ilm, e iewe s would no doub alk abou a “s a -
s udded cas ”. As edi o s, we a e uly g a e ul o all o ou colleagues who ha e
so gene ously accep ed he in i a ion o con ibu e hei expe ise, making i
a ailable ia open access. Like us, hey sha e ou ambi ion o make his book
as accessible as possible so i can be o help o as many colleagues in ansla ion
educa ion as possible as hey lea n o nego ia e he uncha ed e i o y o GenAI.
We a e also g a e ul o Language Science P ess and he edi o s o he “T ansla-
ion and Mul ilingual Na u al Language P ocessing” o gi ing ou book a good
home. This open access olume builds on ou s anding pas publica ions in he
se ies, no leas Kenny’s Machine T ansla ion o E e yone: Empowe ing use s
in he age o a i icial in elligence (2022). This olume also complemen s nicely
Pym and Hao’s How o Augmen Language Skills: Gene a i e AI and Machine
T ansla ion in Language Lea ning and T ansla o T aining (2025) as i ocuses
mo e speci ically on GenAI in ansla o educa ion.
2 Con en s o his book
This book is a icula ed a ound h ee main pa s. Pa 1 explo es he new skills
and compe ences we need o help ou s uden s de elop in he age o GenAI. I
opens wi h Ga y Massey and Mau een Eh ensbe ge -Dow’s de ailed discussion
in Chap e 1 o ansla ion compe ence in he age o GenAI. They make he case
ha GenAI is a collabo a i e echnology ha can be complemen ed by human
agency i human ansla o s b ing hei own, complemen a y skills o he able.
To ha aim, ansla ion p og ammes should o eg ound ex ual, digi al skills and
in e lingual skills when aining s uden s o wo k wi h GenAI. In Chap e 2, E ik
i
In oduc ion
Angelone discusses he impo ance o delibe a e p ac ice in ansla o aining o
help s uden s de elop he compe ence and expe ise hey will need in hei u u e
ca ee s. He hen shows how immedia e, in o ma i e eedback is a key condi ion
o he de elopmen o delibe a e p ac ice and how we can use AI cha bo s in ou
eaching as Mo e Knowledgeable O he s (MKO) ha p o ide s uden s wi h he
equi ed eedback o os e ansla ional expe ise h ough delibe a e p ac ice.
Chap e 3, by Ramón Inglada, a gues ha , in he age o au oma ed ansla ion,
he concep sui abili y should ake p ecedence o e ha o co ec ness. To make
su e ou s uden s a e in a posi ion o e alua e he sui abili y o he ansla ion
solu ions gene a ed by (Gen)AI, he a gues, we should help hem de elop h ee
new skills, namely he skills o selec ion, assessmen and p omp ing, alongside
he adi ional co e skills o ansla ion compe ence. He hen gi es conc e e
examples o how GenAI can be used in he class oom o os e hese new skills
among ainee ansla o s.
In Pa 2, he ocus shi s o he new knowledge we and ou s uden s need o
del e in o in he age o GenAI. This knowledge o en unde pins he new skills
and compe ences men ioned in Pa 1. In Chap e 4, o ins ance, Lynne Bowke
a gues he impo ance o eaching ou ansla ion s uden s abou da a (and, he e-
o e, co po a) as i can help hem make mo e in o med and esponsible use o he
echnology. She ecognises, howe e , ha doing so om an expe - o-expe po-
si ion may no always be mos e ec i e, as i is easy o ge los in a abbi -hole
o in ica e de ails when alking abou AI ne wo ks. Ins ead, she shows h ough
conc e e and compelling examples ha bo owing echniques om science com-
munica ion – whose goal is o make expe communica ion a ailable o non-
expe s – can make ou eaching abou da a much mo e impac ul o s uden s.
This is ollowed by Chap e 5, in which Masa u Yamada discusses p omp ing o
GenAI. Mo e speci ically, he shows h ough de ailed examples he use ulness o
adi ional ansla ion s udies concep s o p omp AI cha bo s, hus opening new
a enues o ansla o educa ion. In an echo o Bowke ’s chap e , Joss Moo kens
and Gökhan Doğ u ackle he all-impo an opic o AI e hics in Chap e 6. They
conside ways o in oduce discussion and e lec ion in he class oom ha max-
imise he impac o his eaching. To ha end, hey ocus in pa icula on case
s udies and he use o mapping and images o os e sys emic hinking a ound
ansla ion and GenAI.
Finally, Pa 3 pu s some lesh on he bones o Pa s 1 and 2 as i e iews some
o he new eaching app oaches adop ed by colleagues since he ad en o GenAI.
I does so by in oducing he eade o a se ies o igne es aken om a a ie y
o ansla ion- ela ed disciplines and con ex s. Chap e 7 opens wi h Senem Öne
ii
JC Pene
Bulu p esen ing he esul s o a esea ch p ojec in which s a and s uden s col-
labo a i ely and expe imen ally explo ed he dynamics o p omp enginee ing
o ansla ion, as well as he new skills human ansla o s may need o de elop
o p omp e ec i ely. In Chap e 8, Ma ia Zimina-Poi o gi es conc e e examples
o ac i i ies o de elop s uden s’ c i ical and e hical use o GenAI as pa o he
ansla ion wo k low. Th ough hese, she con incingly shows how such a use
o GenAI can help augmen human con ibu ions. This is ollowed by Chap e 9,
in which Da id O ego-Ca mona mo es he lens o sub i le aining as ano he
a ea o ansla o educa ion whe e “ he eaching mus shi om eaching spe-
ci ic echnical and ansla ion skills o de eloping adap able p o essionals who
can c i ically engage wi h e ol ing echnologies while main aining high-quali y
s anda ds”. To ha aim, he p esen s us wi h se ies o inno a i e class oom ac i -
i ies ha can be eplica ed o help ou s uden s’ “digi al e lexi i y” in sub i ling.
In Chap e 10, Sonia Vandepi e ocuses on ano he speci ic a ea o ansla o
educa ion, namely ansla ion in o one’s second language (L2 ansla ion). Re-
po ing on he indings o an explo a o y esea ch p ojec in ol ing s uden s o
an L2 ansla ion class on a pos g adua e p og amme, she shows how in eg a -
ing GenAI in L2 ansla ion a he edi ing s age o he p ocess no only enhances
s uden s’ e lexi i y, bu i can also inc ease hei in insic mo i a ion, which is
c ucial in L2 ansla ion aining. In Chap e 11, we mo e o in e p e ing as Saha
O homani and Ne min al Sha man in oduce us o he [AI]ph a p ojec , a sim-
ula ion so wa e c ea ed o enhance in e p e e aining in he MENA (Middle
Eas and No h A ica) egion h ough gami ica ion. Pilo ed in 2023, he key mo-
i a o behind [AI]ph a is o unleash he po en ial o AI o add ess he a - imes
limi ed access o in e p e e educa ion in he MENA egion. Finally, in Chap e
12 we mo e sligh ly away om ansla o educa ion o explo e he po en ial o
emb acing MT in second language (L2) educa ion. In his inal chap e , A sushi
Mizumo o hus in oduces us o he Me acogni i e Resou ce Use F amewo k,
which “posi ions lea ne s as me acogni i e agen s capable o s a egically u iliz-
ing a wide ange o language esou ces, including MT and gene a i e AI ools”.
3 One inal hough
As GenAI con inues o eshape wha i means o ansla e and, he e o e, wha
i means o be a ansla o (educa o ), ou main aim wi h his olume is o help
ellow ansla o educa o s be e unde s and he new skills and compe ences
ha we need o os e in ou eaching, sca olded by new knowledge. We also
hope ha , aking a lea om some o he case s udies in Pa 3, many colleagues
iii
In oduc ion
will be spu ed on o u he explo e wha i means o (lea n o) ansla e in he
age o GenAI in pa ne ship wi h s uden s. This, in u n, could lead o a welcome
inc ease in he numbe o esea ch publica ions on he opic a a ime whe e we
all need o lea n o na iga e ou way h ough his new pa adigm.
Re e ences
ELIS. 2024. ELIS 2024 Eu opean language indus y su ey. Tech. ep. ELIS Re-
sea ch. 1–57. h ps://elis- su ey.o g/wp- con en /uploads/2024/03/ELIS-
2024-Repo .pd .
ELIS. 2025. ELIS 2025 Eu opean language indus y su ey. Tech. ep. ELIS Re-
sea ch. 1–53. h p://elis- su ey.o g/wp- con en /uploads/2025/03/ELIS-
2025_Repo .pd .
Kenny, Do o hy (ed.). 2022. Machine ansla ion o e e yone: Empowe ing use s
in he age o a i icial in elligence. Be lin: Language Science P ess. DOI: 10.5281/
zenodo.6653406.
Koskinen, Kaisa. 2020. T ansla ion and a ec : Essays on s icky a ec s and ans-
la ional a ec i e labou . Ams e dam, Ne he lands: John Benjamins. DOI: 10.
1075/b l.152.
Lambe , Joseph & Callum Walke . 2024. Th i ing o su i ing: Mo i a ion, sa -
is ac ion, and exis en ial sus ainabili y in he ansla ion p o ession. Mikael 17.
89–104. DOI: 10.61200/mikael.136209.
Moo kens, Joss, Andy Way & Séamus Lank o d. 2025. Au oma ing ansla ion.
London, UK: Rou ledge.
O’B ien, Sha on. 2024. Human-cen e ed augmen ed ansla ion: Agains an ag-
onis ic dualisms. Pe spec i es 32(3). 391–406. DOI: 10.1080/ 0907676X.2023.
2247423.
Pene , JC. 2024. Wo king as a p o essional ansla o . London, UK: Rou ledge.
h ps://www. ou ledge.com/Wo king-as-a-P o essional-T ansla o /Pene /p/
book/9781032115573.
Pym, An hony & Yu Hao. 2025. How o augmen language skills. Gene a i e AI
and machine ansla ion in language lea ning and ansla o aining. London,
UK: Rou ledge.
Welocalize. 2023. Emb acing dis up ion in he language se ices indus y. Tech.
ep. Sla o . h ps : / / sla o . com / emb acing - dis up ion - in - he - language -
se ices-indus y/.
ix
