By OECD Event and Communications Co-ordinator Kevin Gillespie
Last week, we hosted the webinar: Using AI to Support Student Creativity: What Do We Know So Far? The webinar allowed participants to hear from people who are deep in the research trenches of AI, education and creativity. And who doesn’t want to find out whether tools like OpenAI are helping students think outside the box or just making it easier to copy-paste their way through school? The webinar is linked to ongoing OECD efforts like fostering creativity and critical thinking and digital innovation in education.
The first speaker was Kim van Broekhoven from Erasmus University who shared research showing that students often rely heavily on AI-generated suggestions when tackling creative tasks. In her study, most ideas from students mirrored those provided by generative AI rather than demonstrating deeper thinking. She’s now working on strategies to help students engage more cognitively, such as using ‘conceptual combination’—a method where learners merge distinct ideas to spark originality. Her work is available here.
Anil Doshi from UCL shared results from a study where participants wrote short stories, some with and some without AI-generated idea prompts. Stories supported by AI were rated more creative but actually showed less diversity. AI seemed to help less creative individuals most, but also made outputs more similar. The paper, ‘Generative AI enhances individual creativity but reduces the collective diversity of novel content,’ is a really interesting read if you have time.
The final speaker, psychologist Todd Lubart from Université Paris Cité discussed four AI usage profiles: co-creation, laziness, plagiarism, and full human creativity. His work emphasises the value of intentional use and the importance of teaching students how to use AI wisely. When pressed to explain the differences about plagiarism and laziness he explained ‘When you plagiarise, you use AI, but you say, ‘I did it.’ When you’re lazy, you just let AI do it and admit it. One is more active, the other is pure outsourcing. But both miss the point of being creative.” I really liked that explanation.
The panelists agreed that AI is particularly strong in idea generation but weaker in the earlier stages of creativity like problem-finding—areas where emotion, context and intent matter. Kim also highlighted the risk of cognitive overload: students can be overwhelmed by too many AI-generated ideas, mistaking volume for originality.
I found it fascinating how some educators are also adapting how they assess creative work. Anil has his students use AI as an auditor rather than a generator. Others ask students to document how they use AI tools, encouraging reflection and transparency. The OECD’s own creativity and critical thinking app (https://www.oecdcericct.com/auth/home ) offers structured guidance in this area if you want to check it out.
The biggest takeaway of many: AI won’t automatically make students more creative—but it can, if it is used with purpose. It might lift some and at the same time limit others. The outcome depends on how it’s introduced, used and perhaps most importantly taught.
As the webinar ended, I was still thinking about that answer Todd gave about the differences between plagiarism and laziness when using AI. And I know what some of you might be thinking, did AI help me write this blog? Well… let’s just say, I’m still learning how to co-create. I’ll let you decide whether it shows 😉
