This LinkedIn message by Barend Last and the reactions it generated triggered us to share our latest activities dedicated to contributing to (language) teachers’ AI literacy development.
In the current EdTech storm (driven by a bigger group of Tech companies than just Apple and Microsoft I experienced in my early days as educator) I noticed that the number of AI-powered applications for Lesson Planning has increased considerably and that the keywords of the marketing approach for these ‘AI assistants’ are time saving and task offloading and promising additional support for other (challenging) aspects of the profession while stressing that the goal is not to replace teachers’ related expertise but to enable them to spend time on what matters: their students.
Focusing on these topics in particular shows that the marketing departments of the Tech Industry have closely followed the reactions of educators during the commotion following the launch of OpenAI in 2022, because it was almost exclusively these aspects that the educational community appeared to appreciate with the least reservations.
Barend’s observation in his LinkedIn post ‘In short, an AI system can be quite useful, but especially because there is a teacher with the right know-how in the loop. Without that person, you can’t really talk about an educational tool, but about a generic script‘ reminded me of a number of Leon Furze’s blog posts about the topic. In the post ‘Lesson planning is a verb: why does tech keep treating it as a noun? he explains why There’s no button for good teaching, which he has also nicely visualised. For the implications for users’ competences see his post ”3 Dimensions of expertise for AI”
This and the very concerns expressed in the paper referred to (Selwyn, Ljungqvist & Sonesson, 2025) about the prompting expertise practitioners need to have to successfully use GenAI LLMs for lesson planning strengthened my expectation that dedicated AI-assisted Lesson Planning applications will be playing a significant role in education (if it were only for the efficient materials generation functionalities they also offer) which motivated us to start exploring what products are available and have to offer through a language and teacher education lens.
To support systematic documentation of the AI-supported Lesson Planning Tools we plan to analyse, we developed a model inspired by McMurry et al. (2016) discussing established CALL software evaluation frameworks published by Hubbard (2006) and Chapelle (2001) from a formal evaluation perspective.
Its current version, based on a first badge of applications, now covers 24 features distributed over 7 categories / aspects.
With more studies being published addressing issues related to Lesson Planning applications and the implications for teachers’ AI literacies and instructional design competences, we also aim to further develop our mapping tool into a resource to be used in pre- and inservice teacher education.
For this we will build on our previous work related to integrating a discipline-specific dimension in generic media and ICT tools, in casu, the WebQuest task template, designed to use new media in an educational setting. This NL project ‘TalenQuest’ (Koenraad, 2005) was selected to be further disseminated across the EU through the project ‘LQuest – Language Quests’, part of ECML’s thematic area ‘New Media in Language Education’, one of the 9 areas of ECML expertise at the core of ECML projects and the work of the Council of Europe.
Once updated after trials in partner teacher education institutions and our ErasmusPlus course ‘Getting better results when Using AI for lesson planning and materials production’ The AI-assisted Lesson Planning Evaluation Framework will become available as OER and the project teams of other projects in ECML’s New Media in Language Education area such as ‘ICT in language teaching and learning’ and its current project ‘AI for language education’ in particular, will be notified.
To conclude, joining the discussion about the apparent AI-limitations in relation to professional development (How can mediocre output ensure quality awareness? And above all: for whom?) a reaction based on our experiences with the use of the a priori task evaluation tool used in the LanguageQuest project, quoting from a previous blogpost about the use of lesson planners in teacher education:
[…] the process of defining the information about learners, preferred methodologies etc. (context engineering) and the critical evaluation of the AI output needed to arrive at results that can function as intended also offer interesting (because dynamic) methodological support to deepen pedagogical content knowledge and development of core competences in teacher education classes.