Now that neural machine translation (NMT) and generative AIArtificial Intelligence (AI) tools have become widely available to the general public—including students—the translation education environment, both in Canada and internationally, has changed radically. Traditional methods of assessing translation competence, centred on the finished product, are inadequate given that there are many pathways to producing the translated text. By shifting to an assessment of the translation process, we can acknowledge the changing environment and devise models that better reflect the wide range of skills being developed by today’s learners, such as self-awareness, creativity and cognitive flexibility. At the same time, such an assessment model recognizes the diversification and expansion of the language-related and intercultural tasks that translators are now performing. This paper examines four pedagogical techniques, based on personal classroom experiences, that can form part of process-oriented translation assessment: (1) commentary writing; (2) portfolios; (3) iterative translation; and (4) reflective reports. Commentary writing, long a staple of translator training, tends to focus on source text analysis and justifying translation decisions, but can take on a more experimental form. Portfolios have potential for translation assessment, though a 2018 classroom experiment with an electronic portfolio system to support self-regulated learning was inconclusive. An iterative approach to translation assessments involves providing partial feedback and repeated submissions and could be adapted to translation using large language models (LLMs). Finally, reflective reports, used extensively in experiential learning, allow learners to draw connections between class concepts and broader workplace and social issues. All these techniques encourage student agency and engagement, and they decrease the weight of the translated text in overall assessment.