The public release of ChatGPT-3.5 at the end of 2022 sparked enormous interest in AI. Within a very short time, the potential applications of AI in academic education, its opportunities, and its risks became central topics in everyday discussions about academic teaching, learning, and research.
Of course, universities have been integrating digital tools into teaching and learning for decades. Just consider the rise and continuous adoption of e-learning since the late 1990s.
However, AI is fundamentally different. The rapid proliferation of AI applications since 2023 has highlighted an unprecedented intertwining of education and technology, profoundly affecting teaching, research, and academic integrity. In response, universities started to invest significant resources to adapt to these developments and to adopt AI in teaching, learning, and research, e.g. by publishing institutional policies on how to embrace AI (cf. MacDonald 2025, Lee et al. 2024).
Many questions and uncertainties about the benefits and risks of AI have been prompted: How should teaching and assessment practices evolve? To what extent are traditional educational models being disrupted - potentially for the better? How can teachers use AI to enhance teaching and learning? How can we ensure the responsible and ethical use of AI?
The possibilities for integrating AI into teaching are extensive, as detailed for generative AI in Chapter ‘4. Designing courses’ of these guidelines.
Regardless of this, one thing is beyond doubt: ignoring AI in teaching and learning in higher education is not an option. Teachers must embrace its potential to enrich, integrate, or fully transform learning environments to ensure that AI’s game-changing opportunities align with pedagogical values and meet the evolving needs of students, and society.
Using AI for teaching & learning requires addressing three fundamental pedagogical questions:
How do learning outcomes need to be reformulated?
How must teaching materials, learning exercises, and assignments be adapted, developed further, or enhanced?
Which methods and criteria are to be applied to assess the work and examinations that students complete with AI?
These questions are directly connected with fundamental issues of academic education which inevitably arise when AI is used. The most important issues include
the use of AI by the principles of academic integrity,
the legally compliant use of AI,
and the ethically responsible use of AI.
While the core principles of academic integrity, legal compliance, and ethical integrity are the same across disciplines, the practical implementation of teaching methods varies considerably between academic faculties and disciplines.
Teachers need AI literacy and skills to support students in developing AI literacy and skills. Within the wide scope of AI, certain skills are essential, such as prompting, using AI in a legally compliant and responsible manner, generating teaching materials or assessments with AI, and critically evaluating AI outputs (cf. Lee 2023). Thus, teachers are encouraged to maintain and expand their roles and skills as architects of meaningful learning environments by embracing professional development in the age of AI (see Chiu 2024). This will enable them to design pedagogically, legally, and ethically sound scenarios in which AI serves as a teaching and learning tool.
For teachers, adopting AI means acquiring skills at four levels:
The ability to use specific AI tools, general and subject-specific, as standard elements of a professional teaching skills portfolio (technical AI skills).
The ability to guide and advise students in the professional, legally compliant and ethically responsible use of AI tools and in critically evaluating outputs (AI teaching skills).
The ability to use, customize, or develop AI tools like chatbots for teaching and learning (AI development skills).
The ability to develop and apply methods for assessing students’ progress, works, and examinations accomplished with AI (AI-related assessment skills).
Building one’s own digital and AI literacy, such as identifying AI's potential and limitations, and devising strategies for its integration into teaching practices, is imperative to leverage it. To this end, the AI-HED project has developed workshops and interactive AI-HED Training Materials for Teachers.
Prompting is the primary method of communicating with generative AI tools. A prompt is an instruction or input provided to an AI system to elicit a specific response. Producing high-quality outputs often requires a structured, step-by-step approach. The practice of giving systematic instructions to generative AI is known as ‘prompt engineering’, which involves specialized techniques and methods to develop and optimize prompts effectively.
Prompting and prompt engineering are essential components of teaching skills in two ways:
First, they are necessary for creating course materials, assignments, exams, and other teaching-related content using AI.
Second, teachers must be able to guide students in using prompts effectively. This entails understanding the types, techniques, and quality criteria of prompts and prompting.
AI companies usually provide valuable prompting and prompt engineering guides (e.g. Anthropic 2023, OpenAI 2023) with explanations of key terms, techniques, practical examples, and more.
For teachers seeking an easy introduction to prompting, a growing number of universities and educational institutions offer useful resources and examples, such as:
Introduction to basic concepts of AI, elementary guidelines, and quickstart resources (metalab at Harvard 2025)
Prompt libraries featuring examples from academic teaching across a wide range of subjects (e.g., Maastricht University 2024, Wharton University n.y.)
Prompt labs for generative AI in university teaching (e.g. AI Campus 2024, available in German)
Prompts and example outputs for various tasks, including developing research questions, writing code, and creating assessments (e.g. University of Edinburgh 2024)
Some applications of AI are well established, while many others are still emerging. As AI is adopted in more and more areas of professional, public, and private life, teachers must stay informed about new tools and concepts. Teachers should also explore ways to integrate AI with other learning technologies and leverage innovative applications in their courses.
Staying up to date requires understanding how AI is being utilized in the occupational fields where students are expected to work. To this end, universities have adopted particular strategies, e.g. carrying out analyses of AI-related qualification requirements and providing this information to teachers and academic experts, for purposes of program and course development.
Irrespective of this, there are various easy ways to keep up to date for teachers:
seek advice from experts from the relevant professional field
analyse which specific AI skills are cited in typical job announcements
take a look at the AI-HED AI Tools Heatmap.
Ultimately, there is one simple recommendation for educators considering AI adoption: Dive in!
The term ‘generative AI’ is arguably the most well-known type of AI, yet it is often mistakenly equated with artificial intelligence as a whole. In academic education, a wide range of AI systems and tools are relevant that not primarily aim at producing new content like generative AI but focus on identifying patterns in data to make predictions or decisions. Here is a selection of types that are used in academic education for different purposes.
Generative AI uses frameworks and models to produce new content, such as text, images, sound, or code. To this end, generative AI uses models like large language models (LLM) to learn patterns, structures, and relationships within the data, which are then used to generate new content. Teachers use generative AI to develop teaching and assessment materials such as presentations, rubrics, educational videos, assignments, and tests.
Predictive AI uses algorithms and architectures to forecast future outcomes by analysing historical trends and correlations. Predictive AI rapidly gains importance in various disciplines, e.g. for business analytics, forecasting disease progression in health care, or protein structure prediction in biology. In higher education, prediction of student retention is a typical field of application of predictive AI.
Assistive AI focuses on models and frameworks designed to support human decision-making or tasks without necessarily generating new content. In higher education, assistive AI is applied to various functions, including research paper analysis, text corrections and improvements, grading assistance, administrative task automation, and enhancing accessibility for students with disabilities.
Adaptive AI adjusts algorithms and models dynamically, learning from new data or environmental changes to improve performance. In higher education, adaptive AI can personalize learning paths by analysing student performance and dynamically tailoring content to individual needs.
Simulative AI employs models and computational architectures to create virtual environments or scenarios that mimic real-world processes for training, testing, or experimentation. In higher education and research, it is used for virtual labs, research simulations, engineering design testing, and modeling of climate models or molecular interactions.