The use of artificial intelligence in education is already a reality for students, teachers, and schools across Europe. With education being highly impacted by AI, we observe both challenges and benefits in its application, highlighting the importance of its responsible use. At Sanoma Learning, we keep up with AI advancements, carefully assessing risks and benefits to responsibly support learning experiences and outcomes.
To dive into the topic, we talked with Oleg Sokolov, Senior Director for Data & AI at Sanoma Learning. In this conversation, part of a series of articles about AI in education, Oleg reflects on the use of AI in learning, the future of education, and our approach to ensuring the ethical use of artificial intelligence. Check out the interview below!
SL: Please introduce yourself and your work at Sanoma Learning related to AI.
Oleg: My name is Oleg, and I am responsible for data, analytics, and AI at Sanoma Learning. I lead Data and AI domain that includes established teams of data engineers and data analysts, as well as a growing team of AI specialists. The goal of the AI team is to define our AI strategy, establish AI center of excellence, promote knowledge sharing across markets, and transition pilots into production.
At Sanoma Learning, we've been testing a lot of different things, such as tools for software productivity, content creation together with our publishers, as well as teacher and student-facing applications and features. Our goal is to improve the experience for schools, teachers, and students. After many successful pilots, we're excited to start implementing AI tools this year, and this requires a lot of work not only in terms of development – but also compliance, communication, governance, data foundation, and various other topics.
My background is in management consulting and AI product development, with experience in setting up and driving analytics and AI transformations, as well as leading teams to build analytics or Generative AI products.
SL: AI is the topic of the moment not just in business, but in society in general, changing the way we work, learn, and consume content. How will AI shape the future of education?
Oleg: Indeed, AI has been developing at an accelerated pace in recent years and will reshape various industries, with education being one of the most impacted. The impact on learning can be significant, focusing on improving the experiences of teachers and students and enhancing learning outcomes. This is an area where we see a lot of interest from our markets. For instance, we have already launched AI solutions, like the survey generator for teachers and immersive reader for students, and we have more AI products in the pilot stage now with very promising initial feedback.
Another area where AI can add value is the creation of educational content, which is at the core of what we do at Sanoma Learning, developing methods for teaching different subjects in various countries. Artificial intelligence, especially generative AI, can increase efficiency and reduce time-to-market, serving as a tool for our publishers who continue to curate content creation and ensure its trustworthiness. We do not believe in purely AI-generated content but see AI as a tool to assist our publishers and authors in the content creation process.
Additionally, AI will improve the experience of our employees, enabling them to bring digital solutions for education to the market faster, particularly with innovations in software development.
SL: Generative AI is here to stay, as other technological tools did before and were incorporated into our lives. So rather than reflecting on whether we should use it or not, the next steps involve discussing how. What is Sanoma Learning’s approach to the responsible use of AI in education?
Oleg: I acknowledge that it's not a question of whether teachers or students use AI, because they already do. Now, the challenge is to enable AI ethically by helping users utilise it without compromising content rights, and ensuring safe learning environments. Teachers and students need trusted tools in terms of the content they provide and in terms of personal data processing – which is a difficult challenge to solve, making it a concern for teachers around Europe.
At Sanoma Learning, our mission is to ensure that, when we bring AI into our solutions, these solutions follow clear ethical principles. When we think about new use cases and develop AI features for our products, a dedicated AI compliance team ensures that all our processes follow these ethical principles for AI.
SL: We understand that the benefits and risks of using AI have to be carefully evaluated. How can we balance innovation and compliance to create value for our teachers and students?
Oleg: First of all, we apply external regulations and internal processes to ensure ethical principles are followed. For example, when we partner with companies and startups that can deliver innovative solutions, we have a clear process and checklist to ensure they comply with regulations and our AI principles. This review process for new vendors and partners is crucial to our strategy.
However, the field is developing very quickly, and it cannot be completely risk-proof. Therefore, it is crucial to stay close to teachers and the broader stakeholder groups, such as parents, government, and schools, ideally co-creating and co-developing solutions. When we introduce new AI features, we need to be transparent about potential risks and assess their impact with input from teachers and other experts – as well as set up some experimental pilot environments where, for example, teachers can safely try different solutions. This approach enables us to innovate and explore without compromising compliance.
SL: Talking about creating value, what opportunities do teachers have to use AI as a tool to facilitate their work? What are the potential benefits to learning outcomes?
Oleg: From our survey among teachers in Europe, the most popular use cases for AI are connected to lesson preparation. Teachers use AI to create specific lesson plans, ideate on class activities, generate additional exercises, and adapt materials to specific needs, such as accessibility or language differences. For example, AI helps in translating parts of the text into students' mother tongues, making lesson preparation more tailored and efficient.
Another use case lies within assessment of homework or written assignments. AI can help teachers assess answers to open-ended questions, saving time by providing detailed explanations and hints on grammar or spelling errors, or flagging irrelevant responses. It is important that AI can provide comments, but it is the teacher who should review feedback and assign a specific grade.
A third promising use case is learning analytics. AI can generate insights based on students' performance, such as identifying students who spend too little time on exercises or who ignore certain learning modules. These insights help teachers recommend better study habits and interventions, hopefully improving learning outcomes.
SL: The EU AI Act regulates AI use across Europe, and this is an important resource that guides our processes internally. Could you tell us more about the Act and how Sanoma ensures all processes and products comply with the Act.
Oleg: This is a very important topic. The EU AI Act is meant to regulate AI systems to ensure they are trustworthy and their use remains human-centric, protecting safety and fundamental rights, particularly for students in our case. The purpose of the EU AI Act is to classify AI use cases and suggest mitigation strategies for each risk category.
Sanoma Learning complies with the EU AI Act across our markets by adopting the terminology and classification of AI systems, as defined by the European AI Commission. Every use case idea at Sanoma Learning goes through a risk assessment process. Depending on the risk classification, there are specific measures we need to undertake, such as training and clear instructions. We have a list of interventions required by the EU AI Act for each group of AI systems. By applying these interventions, we ensure we use AI in an ethical and compliant way.
SL: Still about our internal processes, how do you think AI tools can support our employees?
Oleg: AI can help our employees rebalance their time towards more impactful and interesting tasks by saving time on mundane tasks. This happens in two ways: cross-company AI efficiency tools, like AI for editing emails or summarising meetings, and function-specific assistance, like content generation for marketeers.
SL: When we think about new technologies and trends, we always have people on opposite sides of the scale: those who think AI is the answer for everything and those who are sceptical about it. What would you say to both?
Oleg: So, to people who think AI is just another hype, like blockchain or the metaverse, I would say this wave is different. We already see a lot of applications across industries, such as software development and content creation, that have proven to have a real impact. We're almost three years into publicly accessed generative AI, and if you talk to students and teachers, you quickly realise it is already a habit for many, and it is here to stay.
However, to those who are overly optimistic, I would remind them that what technology can do versus how fast we adopt it are very different things. Especially in areas like education, we need to be extra careful – AI can still hallucinate and has other limitations. We need to learn how to work with it, so it will take time to fully embrace the technology and find the right way to implement it into education of our children. Adoption should be gradual and thoughtful, with teachers as the main facilitators of AI in the learning process.
SL: Every day we see new developments in AI, which come with new opportunities and also challenges. Looking ahead, which developments or future trends should we expect that might affect the education industry or learning process?
Oleg: I think there are a few developments to consider. One trend we see now in AI is the move from systems that resolve specific tasks in a linear way to agentic systems, where different AI agents with various expertise work on several tasks in parallel. In the learning context, you could imagine AI agents helping a teacher prepare for a class. One agent might focus on thinking through the lesson plan and activities, another on coding a small application to demonstrate the laws of physics, and another on learning analytics, analysing how different students understood previous topic and suggesting ways to help those lagging behind.
Another trend is the rise of personal assistants. AI can serve as a helper in almost every type of activity, whether it's grocery planning or summarising meetings. In the learning context, you can imagine a teacher assistant or student assistant as the ultimate way of interacting with technology. Instead of juggling ten tools for different use cases, a teacher could have an assistant to help with any task.
This article is part of a series of texts about the topic to be published in June. Stay tuned to our channels to find out more about AI in education.