EPALE - Electronic Platform for Adult Learning in Europe

Artificial Intelligence has a considerable impact on the landscape of adult learning. Whether this impact is for the better or worse depends greatly on how learning designers go about integrating innovative solutions. In a number of occasions EPALE users have pointed out that design lies at the heart of learners’ success with AI-driven learning offers highlighting the need for a careful approach to design, one that places learner needs in the centre of planning a course or a learning session.

A recent EPALE blog post by Djani Husomanovic quotes a research paper that underlines the importance of design: “AI tools should scaffold thinking, not bypass it. They should prompt reflection, decision-making, and analysis—not just provide ready-made solutions.” Earlier this year, the EPALE and Erasmus+ Conference in Vienna presented participants with a detailed action plan on how to adopt AI in adult education, focusing not only on policy-related and ethical perspectives, but also on the pedagogical dimensions. 

Following up on the blog titled AI for Inclusion: Opportunities for Low-Skilled Learners, this article sets out to take an analytic view on how AI tools may influence designers’ take on the planning process. In doing so, it draws on key learning models that are widely known in the field of adult learning and proposes food for thought with regards to challenges and potentials within applying these models. 

Tentative considerations

Prior to looking at the learning models, it is useful to consider a few overarching questions when it come to AI and learning design. 

Access for success

The width of access to AI apps, large language models, chatbots greatly impacts designers’ capacity to come up with an overall design for learning and actual activities too. Too often trainers experience limited activities and templates unless they are provided a subscription by the training institution. Without subscription to apps, activities can quickly become repetitive or stereotypical. This, apparently, calls for an institutional decision to begin with, which requires training providers to commit to integrating AI into their practice.  

Zero prompting vs. heavy prompting

Once obtaining access to AI tools, designers face an important decision, namely whether to invest in prompting when it comes to planning activities or opting for solutions requiring little to no prompting effort. The latter bears the risk of loosing sight of individual learning needs as prompting provides an excellent opportunity to tailor the tasks to the needs of the learner at hand. Thus, this approach may prove conducive for learning sessions transmitting the same learning content regardless of the learners (e.g. some workplace learning programs, certification courses using quizzes, templates etc). However, prompting AI in great details offer greater agency for trainers to include authentic learning materials (actual resources the learners use in their professional or personal lives), which can increase the relevance of the training to the learner. 

Learner and trainer autonomy

Designers using AI tools need to reflect on the changing role of the learner and trainer. When used with insight, LLM apps offer an unprecedented opportunity in streamlining trainers tasks in some fields of adult learning (e.g. foreign language learning classes) that used to take up much capacity i.e. correction of learner input. Moreover, learners who use LLM tools can validate the truths and authenticity of what they were offered to learn much quicker than in the past. Trainers in this context may take up more of a facilitator role while learners can become more actively involved in research activities, and source critique when it comes to class activities all contributing to increased autonomy of the learner and the trainer.

Re-approaching learning models

Models such as ADDIE, Design Thinking, and Mezirow’s Transformative Learning offer enduring insights into how adults learn best - emphasising autonomy, relevance, experience, reflection, and transformation. This blog now explores how AI intersects with these learning models - highlighting both potentials and risks. These considerations can serve as a starting point for further study and testing by the readers. They are not in any way meant to provide an exhaustive presentation of these models as this would be overly ambitious. 

ADDIE Model

ADDIE (Analysis, Design, Development, Implementation, Evaluation) is a systematic instructional design framework. It ensures educational interventions are tuned to learner needs, clear objectives, and evaluative feedback. The model is widely used in corporate and adult education for its structured and scalable design process, but it has also been widely applied in higher education course design (see one such example here). 

Potentials: AI tools can automate learner needs analysis during the preparation of a course through analytics and can also generate rapid content prototyping. These allow the designer to choose from a wider variety of solutions and learning offers. The evaluation can be supported too by generative AI tools. 

Challenges: The AI tool may oversimplify complex learner context during the Analysis phase, which requires the designer to follow this phase more closely and be ready judge the analysis results. In the evaluation phase AI systems may struggle with assessing soft skills or transformative outcomes.

Design Thinking Approach

Originally appearing in the field of organisational development and later adopted to educational contexts, Design Thinking emphasises human-centred innovation and planning along with iterative testing. The five stages - Empathise, Define, Ideate, Prototype, and Test - allow educators to co-create learning experiences with learners, especially in contexts requiring creativity, innovation, and adaptability.

Potentials: Generative AI can facilitate ideation and rapid prototyping of learning activities to a great extent. When prompted carefully, it can provide designers with a highly context-based learning scenarios, which is essential in Design Thinking. It may also be used to generate sample design solutions that proved useful or successful in the past in similar settings.

Challenges: Overreliance on AI can restrain creative risk-taking of learners. This calls for designers to think about the role of AI in task completion well in advance. Contextualising the learning activity may require the designer some prior experience in prompting precisely.

Transformative Learning Theory (Mezirow)

Mezirow’s Transformative Learning Theory puts forward that adults change their worldviews through critical reflection, often triggered by a disorienting dilemma. Through dialogue and critical discourse, learners challenge assumptions and adopt new perspectives.

Potentials: AI can curate diverse perspectives to challenge biases. Furthermore, generative AI can simulate what-if scenarios for reflective exploration potentially along with the trainer. Using intelligent agents with trainer facilitation can scaffold critical thinking prompts and journaling activities in task completion.

Challenges: True transformation is deeply emotional and interpersonal. This remains the main responsibility of the adult trainer. When taken with little critical reflection on behalf of the designer, algorithmic bias may actually reinforce and not disrupt existing beliefs. Finally, as AI lacks the moral reasoning needed to guide ethical reflection and social justice this is also up to the trainer. 

Concluding thoughts

Artificial Intelligence is redefining how adult learning programs are designed, shifting the focus toward collaborative, empowering, and highly constructive learning experiences. 

With the aid of AI, learning environments can adapt dynamically to individual needs, allowing learners to co-create their own journeys through personalised content, reflective prompts, and real-time feedback. Rather than acting as sole knowledge transmitters, teachers are increasingly becoming facilitators or “travel companions” - guiding learners through rich, interactive, and often self-directed paths. This aligns well with adult learning models such as Transformative Learning, which emphasise autonomy, relevance, and critical reflection. 

While models like ADDIE and Design Thinking offer structure and innovation, AI tools enhance their implementation through automation, adaptive systems, and immersive simulations. However, this evolution also demands educators to critically balance the use of AI with human-centred pedagogical values. When applied thoughtfully, AI can amplify adult learning by nurturing deeper engagement, critical thinking, and personal transformation.