Artificial intelligence meets the classroom: Multilingualism has long been a reality in our schools, and AI can help make it visible, facilitate connections between languages, and support individualized learning. Whether through translation, exploring new vocabulary, or engaging creatively with texts and images, our research investigates how AI tools can enhance teaching by promoting diversity, equity, and engagement.
Artificial intelligence meets the classroom: Multilingualism has long been a reality in our schools, and AI can help make it visible, facilitate connections between languages, and support individualized learning. Whether through translation, exploring new vocabulary, or engaging creatively with texts and images, our research investigates how AI tools can enhance teaching by promoting diversity, equity, and engagement.
Description
Language is the primary medium for learning and participation in education. In recent years, linguistic diversity in classrooms has increased significantly: approximately a quarter of children enter school with multilingual backgrounds. The integration of children who have fled Syria, Afghanistan, Iraq, or Ukraine particularly illustrates the extent to which multilingualism shapes everyday school life. This presents teachers with the challenge of ensuring all students can access lesson content while simultaneously recognizing and leveraging the linguistic resources of children and young people as valuable assets. To address this challenge, targeted preparation in teacher education and a consistent alignment with real classroom conditions are essential. Artificial intelligence offers new opportunities in this context:
translation systems, digital dictionaries, and personalized learning platforms help to overcome language barriers and make learning content accessible to individual learners. At the same time, creative AI tools, such as those for text or image generation, facilitate innovative forms of multilingual and intercultural learning. In this way, AI can enhance educational opportunities for linguistically and culturally diverse learners and support teachers not only in managing multilingualism, but also in actively harnessing it as a resource for successful learning.
Artificial intelligence offers a wide range of opportunities not only to capture linguistic diversity, but also to leverage it as a resource for learning. The following seven approaches illustrate how teachers can use AI to strengthen multilingualism in the classroom:
Selective use of translation tools–
AI-supported translators facilitate access to specialized texts and enable all students to comprehend content more efficiently.
Use of digital dictionaries –
AI-based lexicons support individual vocabulary development and enhance language awareness.
Creation of multilingual glossaries –
AI can automatically compile key terms in multiple languages for classroom use.
Adaptation and simplification of texts –
AI can linguistically simplify complex content, making it accessible to learners at different proficiency levels.
Creative text production –
AI-assisted writing and translation encourages students to compose texts in multiple languages and experiment with style.
Visual stimuli through image AI –
AI-generated images support language learning by serving as conversation starters or vocabulary anchors.
Encouraging multilingual reflection –
AI-supported tools enable language comparison, highlight similarities, and promote metalinguistic awareness.
Providing individual feedback –
AI systems offer feedback on spelling, grammar, and expression across different languages.
„Où est quoi ? – Discovering and Designing Spatial Prepositions through Furniture“
Catergory
Content
Target Group
Multilingual learning group (e.g. Year 6, first year of French)
Subjects
French as the target language, with multilingual elements (German, English, Spanish) from phase 4 onwards
Topic
Using prepositions of place in the context of furniture
Methodological Framework
Ziegesar model:
Demonstration
Comprehension/response
Reproduction
Production
Awareness
Digital Element
AI-assisted image generation using multilingual prompts
1. Demonstration (visual comprehension in the target language: French)
The teacher presents a colourful, detailed image of a room containing various pieces of furniture. While pointing to individual objects, she produces clear and simple sentences in French (approx. 8–12 sentences), for example:
Le lit est à côté de la table.
La lampe est sur la chaise.
La commode est entre le lit et la fenêtre.
Only French is used during this phase. Gestures, facial expressions, and deixis support comprehension. Lexical items and structures are introduced exclusively in the target language.
The aim is to foster initial listening and viewing comprehension. Learners are exposed to linguistic patterns without being required to analyse them grammatically at this stage.
2. Comprehension/Response (Action-oriented phase with French as the input language)
Working in pairs, learners receive picture cards depicting furniture as well as a miniature room (either on paper or in digital form). The teacher provides instructions in French:
Placez la chaise devant la table.
Mettez le tapis sous le lit.
La lampe ? Mettez-la entre la table et le lit.
Learners carry out the instructions. In doing so, they respond to spoken input, connect lexical items with concrete objects and spatial relations, and internalise the meaning of prepositions of place.
3. Reproduction (French – guided speaking through imitation and consolidation)
Learners repeat sentences with fixed structures (chorally or in pairs).
They use sentence patterns such as: Le tapis est sous le lit. La chaise est devant la table.
Furniture images serve as prompts for speaking.
Optionally, a digital whiteboard (featuring movable furniture and language chunks) can support sentence construction.
The target language remains exclusively French. No language comparison is introduced at this stage; structures are first consolidated within the new language system.
4. Production (applying structures through creative prompting and image description)
In this phase, learners actively apply previously acquired structures to produce their own linguistic and visual content. Artificial intelligence serves as a motivating and creative tool.
Part A: Prompting in the target language (French)
Learners use an AI image generator (e.g. DALL·E or a school-accessible alternative such as Fobizz AI or Perchance AI) to create an image of a fictional room with a clear spatial arrangement, based on self-formulated prompts in French.
The objective is to apply spatial prepositions accurately within meaningful sentences.
Example prompts (in French):
Une chambre avec un lit à côté d’une étagère et une lampe sur la table.
Un bureau entre deux chaises, avec un tapis sous la table.
Une pièce moderne où la chaise est devant la fenêtre et le lit est contre le mur.
Un salon avec une table basse sous la télévision et un canapé à gauche de la porte.
Example image (AI-generated with ChatGPT 4.0. on 26.07.2025):
Prompt: Une chambre avec un lit à côté d’une étagère et une lampe sur la table.
Learners plan in advance which objects to include and how to arrange them spatially. Prepositions (sur, sous, devant, derrière, entre, à côté de) are deliberately embedded in grammatically correct and functional contexts, ensuring repetition and consolidation.
Part B: Description and comparison of AI-generated images
After generating the images, learners present them either to the whole class or in small groups. They describe the content in French, for example:
Voici mon image. La chaise est entre le lit et le bureau.
Sur le tapis, il y a une table blanche.
La lampe est derrière le canapé, près de la fenêtre.
The teacher monitors the correct use of prepositions and provides scaffolding where necessary, drawing on language from previous phases.
Alternatively, a whole-class guessing activity can be conducted: two AI-generated images are compared, and learners listen to a French sentence or prompt and decide which image it corresponds to. This reinforces listening comprehension and promotes the accurate use of prepositions in an engaging way.
Part C: Extension – opening up to other languages (optional/differentiation)
Advanced or multilingual learners may formulate prompts in an additional language (e.g. English, Spanish, or German) to promote cross-linguistic transfer.
Example:
French: La table est entre le lit et la fenêtre.
English: The table is between the bed and the window.
Spanish: La mesa está entre la cama y la ventana.
German: Der Tisch steht zwischen dem Bett und dem Fenster.
This fosters intercomprehension and encourages reflection on structural differences (e.g. prepositions, articles, and syntax), thereby preparing learners for the subsequent phase of language awareness.
5. Awareness and rule formation (multilingual language comparison and intercomprehension)
This phase focuses on linguistic comparison, analysis, and metalinguistic reflection. The class collaboratively creates a comparative grid of key spatial prepositions in French, German, English, and Spanish.
Meaning
French
German
English
Spanish
on
sur
auf
on
sobre
under
sous
unter
under
debajo de
in front of
devant
vor
in front of
delante de
behind
derrière
hinter
behind
detrás de
between
entre
zwischen
between
entre
next to / beside
à côté de
neben
next to / beside
al lado de
Guiding questions:
Which prepositions are similar across languages (e.g. entre – entre – between)?
Which languages use multi-word prepositions (e.g. à côté de, al lado de)?
How is the relationship between preposition and noun marked (e.g. articles, “de”)?
Which languages use case marking (e.g. German: unter dem Bett)?
Where is the preposition positioned within the sentence?
Example learner observations:
“In Spanish, de always follows the preposition.”
“In French, many prepositions include de, but the article changes.”
“English sometimes uses multi-word expressions for a single preposition.”
“In German, the article changes depending on the case.”
This phase may be extended through a “language rally,” in which groups compare selected structures across two or three languages and present their findings visually.
Conclusion and outlook
Learners consolidate their knowledge by producing a short description of a room in French, optionally supplemented by a multilingual version. They reflect on the strategies that supported their learning (e.g. visual input, physical activity, language comparison, or AI prompting).
For homework, learners may generate another AI image and provide a French description or compile a mini-dictionary of spatial prepositions in three languages.
