In a Finnish classroom full of children under six earlier this year, a teacher suggested writing a story as a group using a new online tool – artificial intelligence (AI).
With the teacher’s help, the children decided the genre (horror), the story’s plot, and the characters to include.
The teacher wrote all of the children’s suggestions into a prompt for an AI system. It not only generated the text, but also some images to illustrate the horror story – to the delight and surprise of the kids, according to an AI literacy expert who watched the exercise.
The story exercise is one way the Nordic country, which lands at the top of an index tracking resilience to fake news across Europe, is starting to teach its youngest citizens how to interact with AI.
Media literacy creates a public that is “both critically and digitally literate,” making it easier for them to assess information they encounter online, according to the European Digital Media Observatory (EDMO).
For decades, media literacy and critical thinking skills have been ingrained in Finnish schools, from math to history and science courses. But Finland’s education experts say the country is still trying to figure out how to integrate AI into their curricula.
“The students need skills to understand AI and understand how it works,” Nina Penttinen, counsellor of education at the Finnish National Agency for Education, told Euronews Next. “In schools, they need to produce texts without AI”.
Media literacy as a life skill
Finland started teaching its citizens about media literacy in the 1970s, focusing back then on how to interpret radio and TV programmes, experts told Euronews Next.
The most recent curriculum update in 2014 – coincidently, just months after Russia illegally annexed Crimea, prompting a flurry of disinformation in Finland and nearby countries – brought the world of social media and smartphones into the fold.
The curriculum works around a concept called “multiliteracy,” the idea that understanding, evaluating, and analysing different sources of information is a skill for life and not an individual course that children can take.
The curriculum is complemented by approximately 100 different organisations that promote media literacy throughout the country. They also contribute teaching materials to classrooms, according to the Finnish National Audiovisual Institute (KAVI).
In their system, children as young as three start to understand the digital environment by researching some images or sounds that they find funny.
By ages seven or eight, children start to receive guidance from their teachers about whether the information they find online is reliable or not.
A couple years later, students who are nine or 10 begin learning how to put together research, with emphasis placed on which perspectives they are selecting and which they could be leaving out.
Leo Pekkala, KAVI’s deputy director, said teachers might explain in math class how algorithms work and how they’re made.
Ultimately, Penttinen said it’s up to the teachers to decide how to integrate critical thinking into their subjects and lessons and to evaluate whether students are meeting expectations.
Pekkala said their approach appears to be working, citing the limited success of disinformation campaigns in Finland.
Most people “seem to recognise really well” that it is malicious, he said.
“There were certain international conspiracy theories [during the COVID-19 pandemic] that were also spread in Finland, but they never kind of spread very largely and people recognised them rather easily and there was discussion that, yeah, that’s absurd,” Pekkala said.
Literacy skills will help with AI, experts say
Deepfakes are one AI-related challenge in the classroom. The World Intellectual Property Organisation (WIPO) defines deepfakes as videos or images that synthesise media by either superimposing human features onto another body or manipulating sounds to generate a realistic video.
This year, high-profile deepfake scams have targeted US Secretary of State Marco Rubio, Italian defense minister Guido Crosetto, and several celebrities, including Taylor Swift and Joe Rogan, whose voices were used to promote a scam that promised people government funds.
This material is “very, very difficult to separate from real material,” Pekkala said.
The hope is that students will be able to use the skills they learned in school to identify that content in an AI-generated video could potentially be “off”. To confirm that suspicion, students would check another source to verify whether that video was truthful or not.
“The surface technology [of AI] which is developing at high speed, doesn’t take away the need for basic critical understanding of how media works,” Pekkala added.
Children will also learn some signs that a video, picture, or audio clip is fake, for example if it generates a “really emotional reaction,” Penttinen added.
Despite the risks, she added that children need to learn “how AI works and [how] the companies are developing it”.
‘It’s a huge task ahead of us’
Kari Kivinen, an education outreach expert for the European Observatory on Infringements of Intellectual Property Rights (EUIPO), said Finnish teachers are already changing how they work with AI in the classroom.
That could include asking for handwritten assignments in class instead of online essays or specifying that AI can be used for tasks like brainstorming, but not for a final assignment. He cited the teacher’s horror story exercise as a way to introduce young children to AI.
The government introduced some AI guidelines, including recommendations for early education teachers, earlier this year.
The document suggests that teachers disclose how and when they use AI in their own work and to tell their students what errors and biases could come from its use.
Pettinen pointed to some flaws in the guidelines, saying that because they are not baked into the curriculum, schools and teachers may not adopt them. A curriculum-wide review typically happens every 10 years, she added, but it has not yet started.
Kivinen said he is working on a joint AI literacy framework for the European Union and other developed countries, which could provide some more guidance.
The framework, to be published in early 2026, will provide guidelines for how students should use AI, how to communicate that they are using AI, and how to get more reliable results from AI.
Eventually, it aims to eventually measure the AI skills of 15-year-olds in 100 countries, he added. He said the AI literacy framework is “aligned with the Finnish approach”.
“[AI use is] not only a Finnish problem, it’s a problem all over Europe and the world at this moment,” Kivinen said. “The AI tools have been developing so fast that the education systems have not been able to follow it sufficiently so far”.
“It’s a huge task ahead of us”.