Artificial intelligence (AI), specifically with regard to the risks it poses concerning misinformation and disinformation, represents a Pandora's box, but AI and its potential uses are something which should be explored in sectors such as education, International Baccalaureate (IB) director-general Olli-Pekka Heinonen indicated while speaking with The Malta Independent on Sunday.

Heinonen previously served as the director-general of the Finnish National Agency for Education, along with being the former director of Finland's national public broadcaster. He was also formerly Finland's Education Minister, as well as Communication Minister. Heinonen was in Malta to speak on the theme of Unlocking AI in education at the invitation by the non-resident Ambassador of Malta to Estonia and Finland, Kenneth Vella

Considering his long and influential career in education administration, the IB director-general was asked whether he believes AI represents one of the most significant potential revolutions in the education sector, to which he replied that it largely depends on "whether we are capable of using it to solve the challenges that education itself is faced with".

He commented that although there is AI in education, the question is one of equity, and of the quality of teaching and learning. He added that it also raises questions about the relevance of what is learned, how it is learned, and where learning takes place.

"Those are the challenges," he said. "I think it's very obvious that AI can help in all of those challenges." He continued that AI could help with bridging certain learning gaps, and can also help with regard to the administrative workload of teachers, though he remarked that awareness of the risks that AI may pose in education needs to be present.

Heinonen was asked about concerns raised by educators about students using AI in their studies or assignments, with the argument generally being that it detracts from the learning experience.

He said the IB, as an organisation, was one of the first international organisations to advocate against banning AI when the ChatGPT wave started. "We should embrace it," he said, "and we should help educators and students to engage with it."

He said that the IB's educational approach is "very much inquiry-based", one which he said focuses on the capability of asking the right questions and having the curiosity to learn what things are for, in part by looking at the good and bad aspects of something.

"We strongly believe that it should not be banned in education, but that it should be explored," he said with regard to AI.

Asked whether he believes that schools should start teaching about the utility as well as the potential risks of AI as part of their curriculum, Heinonen said that he believes they should. He said that AI will be part of the reality of young people when they are growing up and living their lives, and so they need to be aware of how it functions.

He commented that it is not just a question of understanding AI as a tool to find information or to summarise things, but that it is also about comprehending how society is functioning with AI, and how to be a citizen with the reality of AI, in part by understanding how AI can be used to manipulate for good or bad purposes.

Heinonen was asked if there have been any noted shifts in attitudes or work ethic among educators and students since the use of AI has become more commonplace in education, to which he replied in the affirmative. He added that there have also been certain studies asking students for their opinions, which have found that students who have used AI in practice have more positive views about the possibilities offered by AI.

He commented that he thinks it is the same with educators, but added that, on the other hand, certain risks have also surfaced, such as "the question of equality" and concerns about "ethical bias", as well as "questions of effortless learning".

On the topic of effortless learning, Heinonen said that the concern is that students use AI to get to the right answer and outcome without necessarily understanding how. However, he said that effortless learning does not exist in his opinion, as he said that learning always takes effort and attention.

"The way to incorporate AI into the practices of education is on the way, but I would say that we are just on the very first kilometres of that marathon ahead of us," he stated.

When discussing AI in education, the focus often falls on how students use it, but how could AI affect the workflows of teachers or educational administration more broadly? This question was posed to Heinonen, who said that, in his view, this application of AI is actually the part where the effort should start, as he believes that it is also the part where the risks are the smallest. He continued that AI in this way could be used by teachers for different types of administrative work, such as making study, class, and curriculum plans. He added that AI's potential to help with exam questions and perhaps mark certain exams could also be approaching a point where it should be explored as a supportive tool for teachers.

Having spoken about AI's potential as a tool for educators, Heinonen was asked about whether any concerns have been raised by industry professionals and educators concerning data and privacy issues when involving AI in education.

He replied that, yes, there are many types of risks in this aspect, and added that it is something that schools need to be well aware of, but commented that there are other types of issues which "might be surprising us".

One such issue, which he said he is worried about, is that of "AI coaches and AI friends for young people". He spoke of how there is a young generation which experienced the Covid lockdown and the feeling of isolation that came with it, which he said, in turn, has increased certain difficulties in finding friendships and forming relationships. He commented that when there is an AI solution which is "very human-like and always there for you", then there may be some "danger" that young people move away from human relationships and form a dependency on these AI models. He said that this phenomenon is a "very difficult" and "very dangerous" development.

Continuing on the potential dangers posed by AI, Heinonen was asked whether he thinks that AI presents a potential escalation to the dangers of misinformation and disinformation, particularly when mixing it with education.

He said that, with education, the starting point for a long time has been that facts and knowledge are based on scientific worldview, but that now there is a challenge due to the amount of disinformation and misinformation which "corrupts that basic belief in knowledge". He said that this corruption happens gradually, but that once it takes hold, then no one can disagree or agree because there is no shared solid base to disagree or agree on.

The IB director-general remarked that there is a saying in Finland which essentially goes that a Finnish person does not believe something before they see it with their own eyes, but he said even this no longer functions due to the development of things such as AI video generation.

The Malta Independent on Sunday recently spoke with Professor Alexiei Dingli, who had said that it is feasible for AI video generation to become practically indistinguishable from reality in some short time, and he had said that this potential development poses a lot of risks.

Similarly, Heinonen was asked if he believes that the development of AI video generation was in itself a "Pandora's box" moment, where disinformation may ramp up further.

"Yes, because it's about what is real after all. I think it is a Pandora's box moment because it erodes the trust in societies, and everything we do is based on trust. Societies cannot function without trust," he stated. "If we lose that, then we are talking about entirely different looking societies." He went on to say that societies built on trust stand to lose the most if they fail to establish regulations and effective ways to address the issue.