CIUTI Policy Statement

May 2025


Trans­la­tors and inter­preters face many chal­lenges in today’s rapid­ly chang­ing world. Is there still will­ing­ness to con­tin­ue to build the human cap­i­tal need­ed in many areas to make com­mu­ni­ca­tion avail­able to all? Or do we think that lan­guage tech­nolo­gies have solved the problem?


CIUTI acknowl­edges the role of trans­la­tion and inter­pret­ing tech­nolo­gies in today’s soci­ety. Our mem­ber insti­tu­tions’ cur­ric­u­la include cours­es on these tech­nolo­gies and we cher­ish our strong ties with the lan­guage indus­try that increas­ing­ly makes use of them. As aca­d­e­m­ic insti­tu­tions we are con­tribut­ing to the devel­op­ment of tech­nolo­gies and have a duty to crit­i­cal­ly reflect on their use. Train­ing lan­guage mod­els is the eas­i­est part of the tech­nol­o­gy; train­ing humans to skil­ful­ly and crit­i­cal­ly work with them is the hard­est part. So why are we still invest­ing in the lat­ter? There are a few rea­sons for this:

Lan­guage tech­nolo­gies are now used in every walk of life and are designed to meet users’ needs. How­ev­er, research into the qual­i­ty of machine trans­la­tion and inter­pret­ing is severe­ly lack­ing, par­tic­u­lar­ly giv­en the impor­tance placed on these tech­nolo­gies. The lan­guage tech­nol­o­gy indus­try acknowl­edges that the lan­guage machines pro­duce is not always accu­rate yet qual­i­ty issues have real-life con­se­quences for the peo­ple who depend on trans­la­tion and inter­pret­ing. We train stu­dents to skil­ful­ly inter­vene where the machine goes wrong and we give them the degree of dig­i­tal lit­er­a­cy need­ed to advise stake­hold­ers on the use of technologies.

Whether a machine or a human being, tradut­tore tra­di­tore. Trained inter­preters and trans­la­tors are expect­ed to min­imise errors through thor­ough research, sound revi­sion process­es and a crit­i­cal mind­set. Trained lan­guage mod­els have no such safe­guards. Human super­vi­sion is the only way in which machine trans­la­tion can attain the same lev­el of reli­a­bil­i­ty as human trans­la­tion and inter­pret­ing. Our train­ing pro­grammes seek to equip stu­dents with the tools to make trans­la­tion and inter­pret­ing safe again.

Users place trust in machine trans­la­tions and rely on their out­put for deci­sions they make. When these deci­sions turn out to be wrong, these peo­ple are lit­er­al­ly left stand­ing alone. Providers of ful­ly auto­mat­ed trans­la­tion and inter­pret­ing ser­vices request users of these ser­vices to exempt them from the (legal) con­se­quences of mishaps. This lack of account­abil­i­ty is eas­i­ly under­stood: lan­guage tech­nolo­gies have reached such a lev­el of com­plex­i­ty that the out­put can­not be explained or moti­vat­ed in any cred­i­ble way. Our insti­tu­tions train stu­dents to pro­vide eth­i­cal and explain­able ser­vices, and to moti­vate every deci­sion in the trans­la­tion and inter­pret­ing process.

Lan­guage mod­els feed on vast amounts of data, pro­vid­ed will­ing­ly or unwill­ing­ly by every human being who uses them. When tools are used for trans­la­tion and inter­pret­ing pur­pos­es, the mes­sage to be trans­lat­ed and all the per­son­al infor­ma­tion it con­tains, becomes part of the dataset. In such a con­text, keep­ing infor­ma­tion con­fi­den­tial comes at a price and requires con­sid­er­able upfront invest­ments in equip­ment. Our insti­tu­tions pro­mote con­fi­den­tial­i­ty as a basic right, not as a com­mod­i­ty, and train their stu­dents to use tech­nol­o­gy accordingly.

In many coun­tries legal or reg­u­la­to­ry require­ments are in place that com­pel ser­vice providers to seek lan­guage assis­tance. These require­ments invari­ably use vocab­u­lary stress­ing the human com­po­nent of lan­guage assis­tance. This comes as no sur­prise, as true jus­tice, fair tri­al and sound med­ical treat­ment can only be guar­an­teed by human providers, includ­ing human providers of lan­guage assis­tance. Our pro­grammes train stu­dents to take up these essen­tial tasks in mod­ern societies.

While we acknowl­edge the grow­ing lan­guage tech­nol­o­gy indus­try and the advance­ments being made dai­ly, CIUTI stands behind the need for well-trained trans­la­tors and inter­preters who are able to uti­lize these tools pro­duc­tive­ly to enhance/enable suc­cess­ful com­mu­ni­ca­tion. CIUTI also encour­ages indus­try providers of these lan­guage tech­nolo­gies to con­tin­ue to engage in thought­ful research which care­ful­ly explores the accu­ra­cy, usabil­i­ty and eth­i­cal dimen­sions of lan­guage tech­nolo­gies. Our mem­ber uni­ver­si­ties are eager to part­ner in these explorations…

Found­ed in 1960, CIUTI has devel­oped from a Euro­pean-based organ­i­sa­tion into a world­wide asso­ci­a­tion with 60 insti­tu­tion­al mem­bers in four con­ti­nents. Its mis­sion is to pro­mote qual­i­ty and excel­lence in train­ing and research in trans­la­tion, inter­pret­ing and relat­ed disciplines.