There’s more than one way to learn a language today. Some people thrive in a classic face-to-face course, others prefer the flexibility of online tools they can use whenever it suits them — and more recently, thanks to the rapid development of AI tools, a third option has emerged: learning with AI. As part of UNITE’s learning activities, we wanted to explore what this newest option can actually offer, and it turns out there is now a solid body of research backing up this approach.
As part of the Austrian pilot phase, we ran the AI workshop on two dates, 19 March and 2 April 2026. The first session laid the groundwork, introducing participants to what AI actually is and how it works, in plain, accessible language. The second session then turned hands-on: together with the participants, we worked out practical, everyday ways to use AI tools to improve their German language skills — in a way that felt natural and easy to fit into daily life, rather than like an obligation on top of an already full plate.
And by the way, several recent meta-analyses point to a measurable positive effect of AI chatbots on second-language acquisition, with generative AI tools performing notably better than simple, rule-based chatbots. One review of nearly 50 studies found that AI-supported learning is particularly effective for listening and speaking comprehension — exactly the skills that matter most for everyday interactions at work, at the doctor’s office, or with a landlord.
A few mechanisms help explain why this works so well. First, individualisation and availability: unlike a course with a fixed timetable, an AI language tool adapts to a person’s level, pace, and interests, and is available regardless of time or place. Research on self-determination shows that this kind of autonomy — being able to choose what, when, and how to practise — meaningfully increases both willingness to speak and overall motivation to learn.
Second, short and regular practice sessions consistently outperform occasional long study sessions, a finding known as “spaced practice” and one of the most robust results in learning psychology. The AI workshop leans into this directly, encouraging participants to complete short, regular exercises rather than saving language learning for rare, lengthy sessions.
Third, feedback matters enormously. Corrective feedback is one of the most thoroughly researched areas in second-language acquisition, and studies consistently show a stable, moderate positive effect on language learning — an effect that is often even stronger for people learning a language in a country where it isn’t yet widely spoken around them. An AI tool that gives feedback immediately after every exercise offers a level of responsiveness that a single teacher, working with a group, simply cannot match.
Finally, and perhaps most importantly for people learning a language during an emotionally difficult chapter of their lives: several recent studies show that AI chatbots reduce speaking anxiety, because they offer a judgement-free space to practise without social pressure. Being able to speak repeatedly without fear of embarrassment measurably increases people’s willingness to use a new language at all.
That said, the research is equally clear about the limits of this method. AI systems still show limited emotional sensitivity and contextual understanding, which makes them less suited to tasks requiring empathy or nuanced interpretation. This is exactly why the AI workshop is not designed to replace other language learning opportunities, but to complement them — an additional tool, available any time, within a much broader support system.
__________________________________________________
Co-funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the Social Innovation Initiative. Neither the European Union nor the Social Innovation Initiative can be held responsible for them.
