The rapid proliferation of Artificial Intelligence in modern society has produced a range of transformative effects on economic productivity, social relations, educational practices, and even personal identity formation. At the Human-Centered AI Laboratory of the Faculty of Sociology and Law at Igor Sikorsky Kyiv Polytechnic Institute, we have recognized the profound need to prepare students for an era in which AI is ubiquitous, complex, and capable of reshaping not just future job markets but also moral and civic landscapes. In light of this, UNESCO’s AI Competency Framework for Students has emerged as a global blueprint for empowering learners with critical, ethical, and technical abilities to engage thoughtfully and creatively with AI. The framework emphasizes the importance of providing learners with systematic opportunities to cultivate a human-centered mindset, a keen awareness of ethical issues, the necessary skills to develop practical AI applications, and the ability to design AI systems for genuinely beneficial societal contributions. It also acknowledges that students should be able to navigate these competencies progressively, moving from an elementary awareness stage through more advanced applications and culminating in confident, ethically grounded innovation.
The heart of this framework is found in its matrix of twelve competencies, arranged across four key dimensions and three levels of mastery. The four critical dimensions include the human-centered mindset, the ethics of AI, AI techniques and applications, and AI system design. Each dimension contains three competency blocks, mapped in turn to the three progressive levels identified as Understand, Apply, and Create. Although distinct in description, these competencies intersect in practice, supporting students in refining their ability to reason, create, and collaborate in the evolving AI landscape. The first dimension, human-centered mindset, helps learners focus on the role of human agency and responsibility in AI adoption, while also encouraging them to discover how their civic consciousness and sense of accountability must respond to the power and potential pitfalls of AI systems. The second dimension, the ethics of AI, pushes learners to confront and internalize a set of moral considerations—from privacy and fairness to transparency and societal well-being—so that they can become agents of responsible innovation. The third dimension, AI techniques and applications, equips learners with the core knowledge of AI functionality, data-driven problem solving, and the ability to integrate practical AI tools. Finally, the fourth dimension, AI system design, enables them to move beyond passive use of technology and truly engineer new, beneficial solutions aligned with human and environmental needs.
In order to situate these theoretical constructs in an educational context, it can be helpful to understand how the three progressive levels—Understand, Apply, and Create—naturally support the demands of the four dimensions. At the first level, Understand, learners examine definitions of AI, reflect on current real-world uses of these technologies, and identify potential risks and benefits. At this stage, the competencies that cluster within each dimension are primarily about awareness and comprehension. In the human-centered mindset dimension, for instance, students learn to articulate why people remain accountable for AI’s outcomes, and why it is essential to preserve human decision-making authority when designing or using AI in high-stakes scenarios. In the ethics of AI dimension, students start to analyze what it means to avoid harm, respect privacy, consider fairness, and ensure inclusivity. In AI techniques and applications, they discover the fundamental principles of machine learning and data processing, exploring the differences between various algorithmic approaches such as supervised or unsupervised learning. And in AI system design, they begin with rudimentary exposure to how systems are scoped, how relevant data might be collected, and how a model might be assessed for reliability and safety.
Once learners are comfortable at the Understand stage, they proceed to the Apply stage, which builds deeper, more tangible connections between theory and practice. At this level, the dimension of human-centered mindset requires them to translate their knowledge of human agency into actual decision-making tasks. They might evaluate sample AI tools or case studies by mapping out their social consequences, considering whether these tools preserve user rights or compromise them. They also begin to reflect on the roles of governments, industries, and individuals in shaping the AI landscape, refining their sense of civic responsibility in the era of intelligent machines. In the ethics of AI dimension, the Apply stage compels them to demonstrate an ability to recognize and mitigate biases, examine how AI might generate discrimination if fed skewed data, and see how subtle or overt biases in AI outputs can harm communities. They will also practice abiding by local regulations that protect personal data, especially when conducting tasks such as collecting images or text for training small-scale models. Meanwhile, in AI techniques and applications, learners start coding simple AI-driven products or analyzing open-source AI solutions, thereby honing the practical skills of data wrangling, algorithm selection, or error evaluation. In addition, they investigate how to critically select or reject certain tools based on ethical or contextual considerations. Finally, the AI system design dimension at the Apply level enables them to plan a minimal end-to-end AI architecture: they identify a problem, choose or construct relevant data sets, and test a model’s accuracy. They also compare the model’s performance and reliability and might even propose solutions for improvement. Through these practical experiences, learners understand that engineering AI is not purely technical; it also has to meet social, ethical, and environmental goals.
The final level, Create, epitomizes the highest standard of engagement in each competency dimension, challenging students to become innovative problem solvers and co-creators of new AI solutions. By now, they have moved beyond analyzing or adopting existing AI tools and embark on tailoring, fine-tuning, or even engineering new systems for real-world contexts. For the human-centered mindset dimension, the Create level demands advanced reflection on complex issues, such as how to integrate stakeholder input into new AI designs or how to set up systems of ongoing accountability. Students might lead projects that incorporate direct user feedback loops to ensure that new AI technologies genuinely align with the values and needs of individuals and communities. In the ethics of AI dimension, they learn to embed ethical considerations throughout the AI development cycle, adopting an ethics-by-design approach that covers the earliest conceptualization of a system all the way to final deployment. This might include building models that prioritize fairness by actively adjusting weights or data sampling strategies, or guaranteeing inclusivity by training on linguistically diverse data sets. At this advanced level, learners also have the capacity to critique existing legal frameworks, identify regulatory gaps, and propose new policy directions that would better protect public interests.
In the AI techniques and applications dimension, the Create level drives learners to produce tangible AI-driven artifacts or solutions that address pressing societal or environmental challenges. Rather than confining themselves to standardized tasks, they can conceive novel algorithms or repurpose known ones in unexpected ways. For example, they might employ reinforcement learning to design an eco-friendly city traffic simulation, or build a machine vision tool that helps farmers predict crop yields more sustainably. These emerging ideas reflect a growing ability to creatively tackle complex, real-life problems. And the AI system design dimension at the highest level is no longer about basic training or testing; it becomes a matter of genuine engineering, involving sophisticated forms of problem scoping, data pipeline design, thorough model evaluation, and iteration based on user-centric feedback. By designing AI from the ground up and validating it against ethical and human-centered standards, students learn to see themselves as active players who can shape the future of technology for the common good.
The first cluster of three competencies in the matrix focuses on cultivating a human-centered mindset, which is frequently associated with questions of power, agency, and civic responsibility. This cluster includes an emphasis on preserving human agency throughout the AI pipeline, ensuring that ultimate authority and decision-making capacity rests with people, not machines. The second competency in this cluster, human accountability, underscores the legal and ethical obligations of everyone involved in AI creation and distribution. The third competency, citizenship in the era of AI, encourages learners to critically reflect on issues like digital democracy, the inclusive representation of all stakeholders in AI governance, and the potential for AI to either enhance or undermine social equality. Altogether, these competencies guide students to understand why AI should be aligned with human interests rather than overshadow them.
The second cluster of competencies addresses the ethics of AI, with an emphasis on the nuances of moral judgment and socio-technical responsibility. Embodied ethics refers to the internalization of moral principles so that ethics become reflexive in all AI-related practices, whether analyzing data sets or building new models. Safe and responsible use means protecting privacy, guaranteeing minimal harm, and ensuring that AI outputs do not generate skewed or damaging representations of certain groups. Ethics by design, the final competency in this second dimension, specifically challenges students to integrate fairness, sustainability, and accountability measures from the earliest conception of AI solutions and to carry these principles forward through development, testing, and eventual deployment. By weaving these ethical considerations into every step, learners progress from simply following regulations to proactively shaping AI development with a principled framework.
The third cluster, AI techniques and applications, addresses the essential knowledge and skills that enable students to effectively harness AI technology. The first competency, AI foundations, refers to a knowledge base that includes data structures, algorithm design, basic model training, and an appreciation for how AI systems learn from various data inputs. Application skills, the second block, involves bridging theoretical knowledge to practical uses, requiring students to identify suitable algorithms for different challenges, refine pre-trained models, or even create smaller prototypes that deliver real outcomes, such as chatbots designed for specific tasks. Creating AI tools is the third competency in this dimension, signaling a shift from routine usage of existing systems to a capacity for originality and invention. Students learn to code specialized neural networks, experiment with generative adversarial networks, or test new ways of improving inference speed, thus honing their problem-solving abilities.
Finally, the fourth cluster, AI system design, demonstrates the engineering dimension of AI, where students step into a process that begins with problem scoping: clarifying what exactly an AI system should accomplish, identifying data requirements, and defining metrics of success. Architecture design, the second competency, is about translating the problem scope into a structured approach, laying out the flow of data, choosing relevant machine learning models, deciding on optimization techniques, and handling large-scale computing requirements where feasible. Iteration and feedback loops, the last competency, involve repeated refinement of the entire system through testing, user feedback, and real-world performance metrics. Learners come to understand that AI is never truly finished, but rather an evolving construct that needs persistent maintenance, ethical oversight, and performance optimization.
Integrating these competencies at the Faculty of Sociology and Law of Igor Sikorsky Kyiv Polytechnic Institute poses an exciting challenge and an enormous opportunity to underscore the link between technology and societal well-being. Our interdisciplinary context encourages meaningful synergy between the computational, the legal, and the sociological perspectives on AI. To illustrate, courses in sociology can help students analyze how AI might intensify or mitigate social inequalities, programs in law can emphasize potential frameworks for accountability or the drafting of AI legislation, and studies in psychology can help guide the design of user-friendly and respectful AI interfaces. Additionally, because the UNESCO framework is internationally recognized, aligning local curricula with these twelve competencies can reinforce the global relevance and transferability of our graduates’ expertise.
In practical terms, adopting a spiral approach to teaching can help systematically incorporate these competencies. At early stages, students become familiar with the fundamentals of AI—how data is collected and interpreted, how pattern recognition works, and what it means to design algorithms responsibly. Moving forward, more complex tasks call on them to evaluate case studies concerning biased AI or controversies over facial recognition in public spaces, leading them to refine their ethical stances. Later still, they can undertake group projects that simulate real-world stakeholder interactions, where they design AI solutions aimed at issues such as sustainable development or equitable resource allocation. By iterating through each dimension of the framework multiple times with increasing complexity, learners transform raw knowledge into well-rounded competence.
However, effective implementation requires not just a robust curriculum, but also capable faculty, supportive infrastructure, and the willingness to adopt innovative pedagogical models. Professional development programs for educators are indispensable, as teachers must confidently convey difficult technical concepts while also guiding discussions on ethical and social contexts. Modern computational tools and software must be made accessible, especially for students with disabilities or those in remote areas, ensuring equitable participation in AI-driven courses. Partnerships with industries, governmental bodies, and other educational institutions can amplify real-world relevance, allowing students to see how their newly acquired competencies can address pressing challenges outside the academic environment. The Human-Centered AI Laboratory at the Faculty of Sociology and Law can be a driving force here, offering a collaborative space where students carry out research, test prototypes, and regularly exchange experiences with mentors and peers from different disciplines. By linking teaching and research endeavors, the Lab can anchor the interplay between theoretical knowledge and empirical practice, thereby reinforcing the integrated vision of the UNESCO framework.
STEM students naturally gain particular benefit from mastering these competencies, as they can transition from learning basic AI principles into shaping entire AI ecosystems with a strong sense of responsibility. From the vantage point of law and sociology, though, it is equally important to cultivate the skills that ensure future leaders comprehend and regulate AI in a manner that upholds democratic values and human rights. For instance, law students can develop specialized knowledge on drafting AI legislation that balances innovation with consumer protection, while sociology students delve into the societal ramifications of algorithmic decision-making. Such interdisciplinary synergy can expand the potential of the UNESCO framework, ensuring no dimension remains narrowly confined to a single academic silo. Instead, each is enriched by multiple perspectives, reflecting the real complexity of AI’s expansion into every aspect of modern life.
Beyond the classroom, encouraging students to participate in field-specific projects—such as hackathons that tackle local social problems or volunteer initiatives to build AI systems that cater to underserved populations—fosters a spirit of service, invention, and consciousness about technology’s broader impacts. The design of these projects can be directly aligned with the UNESCO matrix, so students address each aspect in a concrete, hands-on manner. They can conceptualize user-centered solutions, incorporate ethical oversight from the initial planning stages, and systematically gather feedback, all while documenting their iterative design processes. Reflection assignments can be built in to help them articulate why they selected certain data sets, how they mitigated bias, or what rationale guided the ultimate system architecture. By linking practical activities to well-defined competencies, students reinforce their capacity to integrate theoretical knowledge with moral discernment and creative design thinking.
As these changes gain momentum, the resulting shifts in institutional culture can help create a more holistic environment where learning about AI is no longer an isolated technical exercise but a driving force for cross-disciplinary collaboration, reflective practice, and civic engagement. We see a future in which graduates leave our institution not only with the capacity to build sophisticated AI solutions, but also with a strong moral compass and sense of personal accountability that ensures these innovations support human flourishing. This is especially salient in a rapidly evolving era of generative AI, where unforeseen consequences can arise from large language models, advanced image synthesis, or algorithmic content moderation. If each generation of students is taught to evaluate technologies with an eye to ethical requirements, sustainable aims, and the meaningful inclusion of diverse communities, it becomes far more likely that AI’s development will be harnessed for the good of society as a whole.
UNESCO’s AI Competency Framework for Students, with its matrix of twelve interrelated competencies spanning four critical dimensions, provides a comprehensive structure that offers depth, clarity, and adaptability for a wide variety of educational contexts. By following the three progressive levels of mastery—Understand, Apply, and Create—educators and learners alike gain a stepping-stone approach to building knowledge, honing skills, and finally reaching a stage where they can fully engage in meaningful AI innovation. The human-centered mindset dimension ensures that AI remains subordinate to people’s inherent dignity and agency. The ethics of AI dimension grounds every practice in moral responsibility and foresight. The AI techniques and applications dimension offers the technical grounding necessary to think creatively and deploy solutions effectively. The AI system design dimension accelerates learners’ capacity to turn novel ideas into tangible results, ensuring rigorous feedback loops keep them aligned with overarching human and environmental needs. Combining these dimensions into a single, tightly integrated matrix fosters individuals who can grasp the complexities of AI, navigate its ethical quandaries, and strive for a future where technology consistently serves human interests. At the Faculty of Sociology and Law, Igor Sikorsky Kyiv Polytechnic Institute, this alignment with the UNESCO framework paves the way for a robust transformation of our educational programs. It strengthens our commitment to interdisciplinary learning, underpins our emphasis on responsible leadership, and encourages the synergy of technology, society, and law for advancing knowledge and public good. By wholeheartedly embracing these globally recognized competencies, we can guide our students to become active co-creators of the future, ensuring their intellectual growth is matched by a solid moral compass and a firm commitment to sustaining equitable, human-centered progress in the AI era.