AI Has Changed Assessment; Higher Education Has Not Caught Up

AI has already transformed how learners think, solve problems and produce knowledge. Higher education assessment, meanwhile, remains largely unchanged. A clear gap is opening between how students actually work and how they are evaluated.

Students use AI to accelerate learning, iterate rapidly and reach knowledge beyond the formal curriculum. In many cases they are becoming more adaptive and efficient than the systems designed to assess them. Yet they stay discreet about it, largely because current models still penalise or misread AI-assisted work.

A fundamental misalignment

We claim to value critical thinking and real-world readiness, yet we continue to assess controlled, decontextualised outputs.

Frameworks such as learning analytics and computerised adaptive testing show that more dynamic, process-oriented assessment is possible, while automated scoring systems highlight both the scalability and the risks of current approaches. At policy level, the OECD and UNESCO keep calling for competency-based, ethical and transparent systems. Institutions remain slow to adapt.

If this continues, assessment risks becoming ever more performative, while students become more strategic and less transparent.

Final thought: the real issue is not that students are using AI. It is that they may already be learning and evolving faster than the systems meant to measure them.

Beyond Mock Exams: Audio Briefings for the AI Era

Mock exams are not a tenable method of assessment in the age of AI and digitalisation. As an educator in multilingual education and AI-assisted pedagogy, I am challenging the memorisation paradigm that has dominated our classrooms for too long. Mock exams belong to a different era, rooted in the standardisation and industrial paradigm that treated education like assembly-line production. Here is why I reject them, and what I do instead.

Why mock exams fail

Mock exams promote rote cramming and turn students into prisoners of repetition. They kill cognitive dynamism and ignore how multilingual skills, such as contextual fluency and cultural adaptation, demand real-world application rather than one-shot tests from an outdated industrial model.

My approach: progressive mastery, in class and remotely

I favour continuous activities and mini-projects that build competencies situationally. Students apply their multilingual skills through collaborative tasks, simulations and AI-enhanced workflows, developing a deep, layered understanding along the way.

The innovation: NotebookLM audio guides

Today I used NotebookLM to create an audio overview of the final exam structure: an elaborate narrative linking each question to the full network of lectures. Students engage multisensorially, listening, associating content (the module on multilingual practices and ethics, for example) and reflecting. This immersive preparation boosts retention without a single mock exam.

In the AI era of multilingual education, tools like these augment cognition, reduce anxiety and spark motivation. Yet existing structures and policies, especially those plagued by micro-management and control-mania, do not let such illuminating methods be nurtured and implemented.

Rethinking Assessment in the Age of AI

Rethinking assessment in the age of AI is now essential for anyone in higher education or K-12. Assignments and homework, in particular, need re-evaluating.

From Product to Authentic Assessment

Product-based assessment should be reshaped through more authentic methods: case studies, peer-to-peer discussion in class, mind mapping, oral presentations, recorded videos and podcasts. We should also set critical thinking assignments in which the thinking is presented visibly and publicly, for instance in peer-to-peer interaction.

In courses centred on writing, essays, translation or commentary, assessment should be combined with other authentic activities, typically by pairing oral or visual artefacts with traditional techniques. Another innovative approach is to assess the process a student goes through to produce the work. A record of the prompts they used, for example, lets educators see the intellectual investment and reasoning behind the final submission.

What It Will Take

Implementing these changes demands significant pedagogical and philosophical shifts across the education sector, in K-12 and higher education alike. That may mean teaching assistants or tutors to maintain a workable student-to-teacher ratio, or reduced teaching loads so that faculty have the time to create authentic, high-quality learning environments.

The need for political will inside institutions is evident when we consider the realities of higher education today, especially in regions such as the GCC. Faculty development opportunities must be established, and class sizes and faculty workloads reconsidered to fit the new assessment methods, including training in innovative teaching strategies and smaller classes to strengthen engagement and learning outcomes. Adopting a new educational philosophy is imperative if these goals are to be met.

Post-Covid Curriculum: Competency Models, Not Textbook Models

Reviewing or developing programmes in higher education usually follows a highly bureaucratic, product-based pathway: documents, checking, more checking, more documents. Or else benchmarking, which too often means looking at what others did and copying it with slight modifications. That is not curriculum development.

The work institutions avoid

Real pedagogical and curricular engineering means collecting, documenting and saturating relevant data. It is not a pathway that higher education institutions, especially in the public sector, like to take. The same shortcut is followed in accreditation.

What competency-based review is not

There are many models for evaluating and modelling curricula, but the most suitable for the post-Covid context is the one that takes competency as its framework. What we privilege in curriculum evaluation or review is not:

• shifting or renaming course titles and content;
• modifying the outcomes or objectives on paper;
• the religious choice of a specific textbook as the condition for the programme to breathe and develop.

Reshuffle the process

Instead, in our view, the first thing to do is to reverse the order of work:

1. Document and identify the practices of the disciplines at stake in the local context, especially where a corresponding professional practice exists in the local economy and society.
2. Select and classify those practices to develop standards of practice. This information comes from data gathered through interviews, focus groups and surveys.
3. Sit down as a team and align disciplinary knowledge with what was collected from the profession, or from the real-life context of the discipline or content taught.
4. Select and categorise.
5. Only then build the content, activities, pedagogies, training modes and assessment methods, aligned with the newly engineered curriculum.

I hope this clarifies the point.

Competence in Translator Education Is Not as Abstract as It Seems

Accreditation bodies accredit a programme according to very specific guidelines and criteria. Field-specific reviewers scan the programme at the content level and end up deleting, adding or recommending courses. Old-fashioned curriculum evaluation was essentially about changing course content, as if curriculum were a game of chess with courses and recommended textbooks, with perhaps a scan of objectives and outcomes at most.

More Than Moving Courses Around

Not so. Other major factors are at play, not only administrative ones but the "alignment philosophy" and the relevance of content, intended outcomes and any other type of competency (the intended skills, knowledge and abilities to be developed in students). Designing programmes and learning outcomes on the competence model remains a challenge in many higher education contexts. It is a question of funding, policy and a change of mindset.

The challenge is thornier still in the humanities and social sciences. The competence model of programme design and evaluation is most established in scientific and professional programmes, in fields such as agriculture, medicine and engineering, where hands-on skills and abilities can be clearly documented from the profession, then redesigned and modelled into a curriculum that first aims to develop competencies and only then identifies the content that feeds them. That is an entirely different mindset from the usual methods of building, designing and evaluating curricula in higher education.

And in Translator Education?

My first suggestion is to stop stressing only the traditional focus on the product, the text. Let us bring in translator competences as well. After all, who is the agent, producer and deliverer of the outcome, whether translated text, video or audio? The human being: the translator, the subtitler, the interpreter.

In fact, translation competence is becoming an outdated topic in applied translation studies. Let us move on. The PACTE group has been discussing it for years and has run a longitudinal inquiry for two decades. Kiraly, Göpferich and Chesterman have added competencies missing from PACTE's predominantly cognitive model. We should now also be discussing translator competence, instructor and faculty competence in translation, competence in translation curricula, and more.

Yes, Online Translator Education Works

Some still say the online model in education does not hold. This year I had the chance to prove otherwise.

I developed, created digital content for, and taught two of the most sought-after courses in translation technologies, plus one course in advanced interpreting, fully online, for an MA programme at one of the best e-universities in the Gulf region, in partnership with Colorado State University.

I lived the experience from scratch, from programme pre-approval to where we are now: our second semester, with two more to go before the first cohort graduates. The content is highly innovative, some of it new to the region. So far the programme has been a success, and many more students are lined up to register next year. Promising.

New content, new mode

In the Arab region, and especially the Gulf, translation departments are not used to hosting courses such as localisation, online interpreting, fully fledged e-courses in translation technologies, revision or project management. These are top-tier innovations in graduate translation curricula. For such courses, students log into Blackboard Collaborate (the chosen LMS) every week, attend a live and interactive lecture, and then complete varied assessments of their understanding.

Diversified assessment is the key element that makes the difference, and it is why the synchronous online model deserves credit. Believe me, learning, teaching and transformation are taking place. Triangulating assessment tools; student-content, student-instructor, student-student and instructor-content interaction online; and delivering assessment to strict deadlines are clear signs of educational practice that creates impact.

So why the doubt?

Why should these learning and teaching experiences not be considered viable? On what basis? Perhaps an ideological, self-serving and narrow interpretation of what e-learning means. With COVID upon us, it cannot be denied that a well-designed online exercise or test should count. Institutional support should also be offered so that instructors can develop abilities and pedagogies specific to online environments.

One last question: what are we assessing, content or understanding? It is time to move beyond assessing the product (content) and focus on assessing the level of understanding, deep or surface (Bart, 1995).

Covid-19 Has Exposed Our Neglect of Formative Assessment

Formative assessment has been badly used and purposefully ignored by education institutions. The video linked below explains why it matters; here is my view on why it has been sidelined.

Why Institutions Avoid It

Perhaps because it is tough on instructors: it needs hard work and a long commitment to continuous feedback, which can be torture for teachers with large classes. For administrators and educational managers, it is not cost-effective, because it forces them to limit class sizes, even as they boast about student-centred approaches. The contradiction is plain: how can an instructor give full commitment to a hundred students through active pedagogies and, by extension, formative assessment? It is time for higher education bodies to rethink their assessment tools and methods and humanise the educational experience for the next generation.

What Needs to Change

The hype and fear surrounding "common exams", "mid-terms" and "finals" needs to stop. Instead, we need to train and recruit responsible, well-rounded faculty with sound ethical and professional stamina. A PhD in some content area will never be a fair or sufficient standard on its own. Genuine professional, pedagogical and curriculum know-what and know-how are key. Only those with pedagogical know-how and a real feel for professional practice can design innovative, practical and well-monitored assessment tools.

Higher education policy must change too:

• reduce class sizes;
• rethink recruitment standards;
• discard the whimsical, amateurish and self-serving political considerations that creep into recruitment;
• treat faculty as professionals, not as puppets or mere transmitters of textbook content who can teach "whatever course" and handle "whatever class size".

Here is the link: Assessing Student Learning Online (Johns Hopkins Engineering for Professionals)

Dictionaries in Translation Exams: Why the Debate Is Misplaced

Many foreign language and English departments in the Arab context, though not translation departments or schools, still treat the use of dictionaries in classroom translation tests as a hot topic. In contexts with a long tradition of teaching translation, the topic does not exist, even where translation is a minor (see the minor programme in translation practice at McGill University, Canada). They went through this and binned it decades ago.

It Is All About Perception

The debate boils down to how our colleagues in neighbouring disciplines, especially EFL, perceive translation as an academic activity and as a practice. The argument survives as long as students see translation as a tool for improving their foreign language proficiency. The worst case is when these colleagues insist in committees that exams be run without dictionaries, as if a translation test were an exercise in text comprehension or vocabulary. Not quite. Translation must be conceived as a profession and a practice. A practitioner needs tools to work and think, and so does the translator, who needs them to solve linguistic, cultural, communicative and ethical problems on the way to rendering a text in another language.

Assess the Process, Not Just the Word

Dictionary use itself, monolingual or bilingual, deserves scrutiny. Do students use the tool efficiently? Do they finish on time? Information mining, searching, speed and efficiency are skills that belong in assessment; they are key professional and lifelong abilities. Marking a student on whether they found the correct meaning of a word is not a sufficient standard of performance. We are measuring the process as well as the product: the use of a tool to choose the appropriate term and solve the problem, and the understanding of a text not for its own sake, as in language learning, but in order to transform and use it for communicative and social purposes.

Choosing the text is the other critical step, and it belongs to the faculty member teaching the course. The instructor needs to select strategic material containing the patterns and structures students must reformulate to demonstrate their mastery of the mechanics of translation and transfer. Whether they used the correct word comes second. Here is where the real problem lies: can you choose texts with the variables that measure the targeted skills, knowledge and abilities? Let us not hide behind the curtain of banning dictionaries. If the material is chosen strategically, dictionaries will not harm the outcome we are after: communicating and reformulating structures and ideas fluently, idiomatically, contextually and ingeniously. The vocabulary-matching game comes second.

Translators Communicate Information, Not Words

The aim is to see whether students can transfer and transform structures at the textual and generic levels to produce healthy, communicative, pragmatic texts with a smooth flow of information. Translators communicate structured information primarily, not just words or correct vocabulary. Much of the traditional translation studies literature ignored the term 'information', deliberately or not, preferring 'source versus target text' or 'transfer of meaning', in other words anything explicable with concepts from linguistics, language studies, or comparative literature and rhetoric. And often students do not find adequate equivalents in the dictionary anyway, so they fall back on context, audience, purpose and other variables.

A Thin Line, and a Teachable Habit

Making translation students sit tests without a dictionary is untenable. We need to draw a thin line between learning a language and being able to translate. Students sometimes produce bad translations even with dictionaries. The trick they are taught is: first do the terminology and glossary, then start reformulating the text. Yet they apply the method they use in language learning, looking up the words in each sentence and translating that sentence at the same time, which leads to literal translations and strict adherence to source-text structure. This can be remedied in the classroom through effective, engaged instruction, monitoring and mentoring by the trainer.

Finally, think of an engineer, architect or mechanic. When they spot an anomaly in their material, the equivalent of a difficult word in a text, what do they do? They use materials and tools. The same applies to the language service provider, translator or interpreter.

After graduating, I sat translation tests at many local, regional and international organisations. I always took a bag of paper-based glossaries and all sorts of dictionaries into the exam. So did every other candidate. Full stop. Outside the EFL universe, it is normal practice.

The Standardised Exam Syndrome: A University Practice to Rethink

Standardised, collective exams have no place in a university. They are a secondary-school habit that many higher education institutions in the Arab world still cling to, and they undermine learning, faculty and the mission of the university itself.

A borrowed practice

Research in the educational sciences shows that standardised, collective examinations at university are borrowed from assessment and curricular practices at secondary level. European universities used them until the late 1970s. Yet many university administrators across departments in the Arab world still import the testing practices of secondary schools, preparatory years and language centres into a university context. Because of the 'mass' or multiple-section syndrome, administration prefers to run exams collectively.

This practice is didactically and pedagogically unsound, even etymologically, and it leads to poor learning outcomes. Micro-learning is far preferable at this level of education. It leads to completion of the learning task, because students can follow the stream of thought around a concept, whether in literature or any other declarative knowledge, in the techniques and patterns of a translation, or in the application of concepts in applied or theoretical linguistics. When other instructors are brought in to standardise the exam, students cannot produce the same quality of outcome, because the close continuum of thought and reasoning is broken. Final exams are a different story.

Three disciplines, one exam

Take the English departments of the Arab region, or an entire faculty of languages. Each is inhabited by three mutually unintelligible disciplines, linguistics, translation and literature, plus, of course, the language institute or centre. Despite this, administration runs the exam the same way for all of them. Can you set questions in drama, research methods, technical or general translation, or interpreting techniques the same way you assess language or preparatory-year students? The knowledge frameworks, pedagogies and practices (professional practices in the case of translation and interpreting) are simply not the same. So why apply one model of assessment, usually the product of an industrial, factory-line model, to every class?

Students on a production line

Students are treated like a production line. They are placed in lanes and rows. We are not looking at their learning; we are waiting for them to reach the end of the line and hand in their paper, not to demonstrate their learning and performance in different ways, but to submit what they memorised by heart. The mere thought of the mass exam hall frightens a student into devouring the content, or not. It is not one hour that reveals the real student and the evidence of learning, but a series of formative and summative assessments over the entire semester. The mid-term, and even the final, should not always be unified. It depends on many variables, starting with the nature of the discipline.

Assessment is discipline-specific

What is lacking is knowledge and skill in designing university assessment. At university, assessment is discipline-specific. Translation and interpreting have a professional vocation, so assessment should target skills, knowledge, behaviour and values, not merely textbook content. We cannot evaluate students only on what is in their textbooks and then administer standard exams that annihilate whatever remains of their learning experience.

Education needs proper investment: the recruitment of competent faculty with pedagogical knowledge of their discipline (and, where possible, of the profession) who can take on the burden of designing sound assessment methods that lead to lasting, lifelong outcomes. Certainly not the mass, line and lane model.

Faculty colleagues need to do this work, at least for the first and second mid-terms and perhaps the final as well in translation and interpreting. If we claim that student translators will operate confidently in the profession upon graduation, then we must prepare activities, content and assessment methods that improve students' performance in knowledge, skills, behaviour and values. The standardised exam is not that kind of assessment, I am afraid.

Good practice versus poor practice

In good practice worldwide, especially in higher education, the hot topics are discipline-specific pedagogies, didactics and assessment methods. We cannot mix everything in one pot; these areas of knowledge are fundamentally different. In poor practice, as in some universities in the Arab context, we still believe that to teach at university, in an English department or elsewhere, all we need is:

• Textbook X
• Lecture up to chapters x, y and z, then run mid-term 1, mid-term 2 and the finals en masse, with coordinators collecting questions from faculty
• Call in the army of invigilators to police the 'villain' students
• Call every academic an 'instructor' during exam time, diminishing their status and professional standing, as if they were chess pieces to be moved wherever and whenever administration pleases. Full, associate and assistant professors are rushed out to do four or five invigilations in a miserable, stressful, unhealthy, unethical and non-pedagogical environment. There are universities where this happens.

What this does

Approaching exams in this way leads to two outcomes:

• It kills the mission of the university. Such assessment works against student success and against the lifelong, durable, social and interpersonal skills that the university's own mission statement and outcomes proclaim. Institutional inconsistency is unacceptable.
• It demeans faculty. Academics are reduced to 'guides', 'servants', teachers of textbooks and exam markers, the same model used in primary and secondary curricula. Administrators forget that faculty, especially those trained at good international institutions, gained academic freedom and the ability to produce knowledge through long doctoral training. Those powers are scrapped once unhealthy administrative and assessment practices are imposed unscrupulously. Faculty barely manage a decent empirical article a year under such conditions. They become mere textbook teachers, there to make life easier for administrators who prefer the factory, army and counter model of evaluation inherited from early twentieth-century curriculum practices in secondary, or perhaps primary, schools.

Takeaway: assessment at university must be designed by the discipline, for the discipline. The standardised mass exam is a relic, and it costs us students, faculty and the mission we claim to serve.

Two Experiments in Translator Training: Case Studies and Project Work

Training and educating translators differently means innovating in course design, course development and classroom interventions, whether face-to-face or online. Here are two experiences from my own teaching, in two very different settings and modes: an online synchronous course for practising translators, and a face-to-face course for undergraduates in Oman.

Experience 1: Case Studies in a Synchronous Online Course

In 2012 I took a challenging step: teaching and facilitating an online synchronous course on one aspect of 21st-century professional translation practice, cloud sourcing. It was the first time I had taught such content in Arabic, and a real challenge in terms of concepts and terminology, since the metalanguage for this content is still emerging and far from stable. That is true even in countries with a long tradition of translator training, let alone the Arab context, where university translation teaching is still, in many places, a new tradition with much work to do in pedagogy and in the discipline at large.

I divided my intervention into two phases. First, an online PowerPoint lecture, delivered through a virtual interface so that students could see the slides, in three major components. At the end of each component I paused for questions, which students asked via the chat box or the audio-visual channel. Then I moved to the practical part, using case studies tied to the theoretical content: for instance, a client–translator or project manager–vendor negotiation over tasks, technical issues and pricing. I looked for cases that matched the skills and concepts I wanted students to appropriate and interact with.

The participants were already practising freelancers, and some owned their own agencies, which made the second part of the course very rich. I drew on educational constructs to guide me: learning styles, learning strategies for adults (andragogy) and studies in psycho-education (motivation). Professionally, all of us, myself included, benefited from the experiences and practices described by participants living in Australia, the USA, Europe and the Arab world. Each participant resolved the problem, or proposed a solution, according to the local practices of his or her country. This intercultural dimension enriched our global understanding of the profession.

One more design choice: the lecture was in Arabic, while the practical part was in English. Many participants have Arabic as a passive language: they understand it well, but not at the oral level. As adult learners of various nationalities, meeting for the first time and online, I judged it better to let them work in the language in which they could express themselves at ease.

Experience 2: Project-Based Learning in Oman

The second innovation was in a face-to-face classroom. In 2013 I taught translation and interpreting (community interpreting) to female undergraduates at an Omani university. I had four courses: modern theories of translation, Arabic–English practical translation, introduction to interpreting, and business translation.

For the practical English–Arabic course, taken by final-year students, I decided to apply Kiraly's project-based approach. When I spent nearly an hour explaining what project-based work is, many students felt uncomfortable: the practice was new to them and very different from the traditional (collective and pastoral) methods they were used to in year 3, for example. I went ahead for two reasons. Class size: there were 22 students, and it would have been difficult with the 30 or 35 I had in other classes. And maturity: as final-year students, their level of maturity and language mastery helped, in a way, to make the approach work.

The Process

It took me a while to work out what activities would motivate students and make them see the relevance to their future careers. So I checked, and checked again. I found that the university's HR department published its employee manual in English only. No Arabic version. Strange, isn't it? I visited the HR manager and explained the initiative. She liked it, naturally: a free Arabic manual is not a bad deal. But I defended my students by proposing an officially stamped certificate of recognition they could add to their portfolios and CVs. She agreed.

The following week I told the students how they would be compensated. They were buzzing. They liked it even more when I said the activity would count as formative assessment. I told them I would note observations during every class session of the four-week project, on each candidate and each group: how each member put in effort and committed herself to the project, and the dynamics of each group.

I divided the class into seven groups of three. Each student took on the roles of terminologist, translator and reviser, and students checked their colleagues' performance (peer assessment). To my surprise, this provoked a range of negative attitudes. Some did not like having their translations criticised (a very important cultural variable for my ongoing research). Some who did not want to complain in front of their colleagues came to my office to do so. In the next class I explained further how they should adjust their attitudes: this was a project, and they were collaborating for one purpose, to hand it in at its best, on time, and get paid for it (the certificate). It was a first step toward the professional dimension of the activity, a transition phase, and such reactions are part of it.

What the Students Said

At the end of the project I gave them a survey with one question: describe your experience. No names, ten minutes, and I left the room. Reading the replies at home, I was struck by the change this activity had produced, especially in their attitudes and in how they conceive of translation: as a profession, not as an exercise to improve their L2 competence, which has been, and still is, the tradition in this context. I take this to be what Kiraly (1995, 2000) calls the development of the self-concept. Two students expressed dissatisfaction: translation was not what I had shown them, they said; they had enrolled to improve their language and become language teachers after graduation.

Takeaway

These are two true experiences I wanted to share with my community, in the hope that colleagues will engage in more innovative initiatives of their own. But such innovations must be backed by insights from the educational sciences: concepts from higher education pedagogy and curriculum studies.

Do Online Students Cheat More Often?

An infographic asking whether online students cheat more often than their face-to-face peers – a question every online or blended programme has to face.

Do Online Students Cheat More Often? (e-Learning Infographics)