Developing teachers’ AI literacy through practical teaching tasks

Using generative AI in teaching requires educators to evaluate its suggestions against learning objectives, student needs and assessment requirements. Teachers need opportunities to practice with real teaching tasks and judge the quality of the output.

The Erasmus+ PATHFINDER project (LAB 2026) supports this work through SAGE, which provides practical guidance, and GUARD, which offers structured learning labs. LAB University of Applied Sciences has contributed through workshops, teaching materials, ideas for tasks and testing.

In autumn 2024, LAB organised the “AI for Teaching and Learning” programme. Sixteen teachers participated across the Lahti and Lappeenranta campuses. Five sessions combined discussion, practical exercises and opportunities to try AI tools.

Topics included ethical questions, prompting, lesson planning, assessment and guiding students’ AI use. Participant feedback highlighted the value of practical examples and hands-on activities. Teachers also wanted more discipline-specific examples, practice time and follow-up training. Managing student AI use and academic integrity remained important concerns (Bogdanova 2025).

LAB provided materials informed by the workshops and feedback to the project’s lead partner, who incorporated them into SAGE. This brought experience from teaching practice into the development process.

From guidance to practice

SAGE, Strategies for AI-Guided Education, covers generative AI basics, prompting, teaching tasks and supporting students’ learning with AI. Educators can select guidance for their needs, including planning lessons, creating assessments and setting AI-use rules (PATHFINDER 2026a). The platform makes it easy to find practical information on a wide range of questions related to the use of AI in teaching and learning.

GUARD complements this guidance with five labs: planning a class activity, creating an assessment, designing a rubric, revising a curriculum and revising an assignment for AI use. The labs support prompt formulation, critical evaluation and refinement of AI-generated suggestions (PATHFINDER 2026b). Each lab guides users through a series of questions and generates a prompt that can be further edited and copied into the LLM of their choice. This process helps teachers consider relevant details systematically and formulate more effective prompts.

One practical example of how SAGE and GUARD can be incorporated into teaching attracted particular interest from educators. Many students already use AI in their studies, raising an important question: how can assignments accommodate AI use while ensuring that students remain responsible for the thinking required to achieve the intended learning outcomes? AI can be used to augment their thinking and learning and strengthen critical thinking, rather than relying on it primarily for cognitive offloading.

GUARD’s “Revise an Assignment for AI Use” lab helps teachers explore this question. They can upload an existing assignment and work through a guided revision to clarify AI-use expectations and preserve meaningful assessment.

The learning objective is the starting point. If students need to analyse, evaluate or reflect, the task should require them to exercise those skills. AI could be required, permitted for supporting tasks or restricted where independent work is essential.

Revision can make evidence of learning more explicit. Students might connect course content to their own context, justify decisions or explain how they would apply their learning. AI may support the process, but the assessed work should demonstrate their understanding and reasoning. Describing AI use alone does not establish that learning has taken place.

Teachers should check whether the revision preserves the purpose, sets a reasonable workload and provides fair expectations for students with different needs and access to tools. SAGE can support this review, while GUARD provides a structured opportunity to develop and question a revision. The educator’s judgment remains essential.