Artificial Intelligence in Learning & Development
Artificial intelligence has the potential to fundamentally change the way we learn, develop, and share knowledge. Our learning systems are constantly evolving, and AI offers new opportunities, such as personalizing the learning process. Adaptive learning systems can tailor content, learning schedules, and timelines to individual needs. In this article, we will explore what is already possible today and what cognitive learning systems might offer in the future.
AI can also support Learning & Development professionals in planning, implementing, and evaluating learning and development initiatives. What opportunities does the use of AI offer in corporate learning? How is AI changing the work in Learning & Development? What prerequisites must be in place to leverage AI effectively in corporate learning?
The integration of artificial intelligence (AI) into corporate learning promises to fundamentally change how we learn, develop, and share knowledge. The opportunities that arise and the challenges that must be overcome have the potential to significantly transform our learning landscape.
The use of AI in corporate learning is not new. Due to the hype around generative AI, the topic has been more present than ever for some time now. AI will fundamentally change the way knowledge is acquired.
The potential is diverse, ranging from the creation and optimization of learning content to course recommendations for acquiring skills and personalized learning support. For example, chatbots can be used to serve as virtual learning partners for employees.
However, even though AI has been regarded as an important tool in corporate learning for about 10 years and 90 percent of education experts from the Cornelsen Education Index see AI systems as either an aid or an innovation driver, a Bitkom survey shows that only 3 percent of companies use generative AI centrally. The understanding of suitable use cases varies significantly across different organizations.
Currently, we stand between opportunities and challenges: AI offers enormous potential as a supportive tool, such as in process automation. This does not mean that human skills and jobs in corporate learning become obsolete; rather, it enables a new, symbiotic collaboration between humans and technology.
Successfully implementing AI in corporate learning requires a balance between technology, organizational processes, and the people who work with them. Crucial is the overall transformation of an organization. Without adjustments in processes and employee training, AI integration cannot achieve the desired results. Therefore, it is essential for companies to invest not only in technology and processes but also in employee training. Only then will a successful and legally compliant implementation of AI in corporate learning be possible.
Rule-Based, Data-Based, Cognitive? The Development Towards Adaptive Learning Systems
Personalized learning systems in corporate learning have been discussed since the 1980s and 1990s. Learners could often choose from multiple learning paths or modify avatars in gamified environments. However, this was not personalization of the learning process in the modern sense and was not based on AI. Today, these systems are called rule-based systems.
Current developments are dominated by data-based learning systems. These systems collect user data and make decisions at predetermined nodes based on the user's data. An example of this is the language learning app Duolingo.
The future vision shaped by AI goes beyond this and consists of cognitive learning systems (deep learning systems). In this vision, teachers are replaced by AI systems that are indistinguishable from humans in their language and behavior. These systems are based on human-machine interaction (HMI) and will be able to create training entirely on their own based on data.
Among the existing systems, large language models (LLMs) like ChatGPT or enterprise-focused AI like WatsonX come closest to this vision, but they are still far from it.
Today, we find ourselves between data-based and cognitive learning systems. In the literature, these possibilities are summarized under the term adaptive learning systems. These include all systems that independently evaluate learning and user data and enable personalized learning for users. This personalization goes far beyond binary data, incorporating personal and statistical data. Adaptive learning systems use AI and data analytics to maximize learning success through individualized learning experiences. Initial pilot projects are already in development and use of such AI learning systems.
Adaptive Learning Systems: The Future of Learning
What is Already Possible Today?
The market for adaptive learning systems is currently growing by more than 20 percent per year. After Asia, Europe is the second largest growth area worldwide. North America already has a very high level of penetration.
Adaptive learning systems not only offer users a personalized environment and learning experience that takes multiple factors into account, but they also learn from their users and automatically improve based on collected data.
However, the general practice does not yet match the possibilities and the state of research. While many existing systems rely on pre-existing datasets, it is also important to link these with additional data. For example, the mere number of clicks on a video sequence in a learning system does not indicate why learners clicked on that sequence. So far, many data-based learning systems do not consider a lack of prior education or different goals.
Valuable adaptive learning systems give users the opportunity to adapt the learning environment based on personality, individual goals, and many other factors. One important factor is considering prior knowledge to narrow down suitable learning content.
In summary, good adaptive learning systems can:
- Configure learning goals, content, and pace
- Adapt content to personality and personal characteristics
- Adjust learning paths to the current competence level and offer individual support when needed
- Customize learning and schedules to user needs
What Will Be Possible in the Future?
It is still unclear when the vision of cognitive learning systems will be achieved. Adaptive learning systems will continue to evolve rapidly in the coming years, driven especially by the data collected, promoting greater personalization. Learning systems will respond to the experiences of their users and, for example, independently recognize when a generational change requires different personalizations.
Imagine a teacher who knows their students with all their characteristics and adjusts to each one 100 percent.
Personalized learning experiences and cognitive systems that understand the strengths, weaknesses, and preferences of learners and respond to their needs will continue to develop and could become the norm.
The consent of learners to store their data and be open to personalization will be a key factor in the further development of adaptive learning systems. By jointly processing personal information and technical data from learning systems, algorithms can offer more efficient learning and more personalized experiences to future users.
However, the analysis and processing of these data also pose risks. Therefore, we will need to address the question in the future of where we should remain open to technology and where we should set boundaries.
AI is revolutionizing the learning concepts we know. New tools for text summarization, interpretation, speech-to-text, and text-to-speech can be realized. This increases personalization, adaptability, and inclusive use of learning content. In addition, AI can replace passive learning approaches, such as mere information intake, with more active practices like retrieval, reflection, or linking information.
In the department responsible for Learning & Development (L&D), staff can analyze learners' behavior and progress in real-time through integrated AI in learning systems. Based on this data, the system can directly adjust the learning content. This strengthens learners' motivation and engagement and improves the overall efficiency of the learning process. Various AI technologies can also capture individual learning preferences, knowledge levels, and progress. With this data, personalized learning paths can be created that meet the individual needs of learners. By analyzing employee data, generative AI tools can provide recommendations for necessary training or pose more appropriate reflection questions, enabling employees to develop more targeted and effective skills without new prepackaged programs.
Which Factors Must Be Considered When Using AI?
To successfully use AI in corporate learning, we need to clarify which prerequisites and competencies learners need to interact with AI. This should be distinguished across three levels: the level of the companies (organization), the level of the responsible managers and leaders, and the level of the learners.
1. Level: The Companies (Organization)
More and more companies recognize the importance of L&D, and the field is now often represented at the board level. This demonstrates the strategic relevance of L&D in the context of disruptive changes. A first prerequisite is defining a company-wide learning strategy. This strategy sets out where a company and its core competencies are heading, motivates the workforce to complement individual learning objectives with AI, and establishes AI use in the L&D field.
The second prerequisite is an actively lived learning culture that promotes innovation and experimentation. This includes a certain level of uncertainty due to continuous technological development and experimenting and failing in collaboration with AI. Companies must lay the groundwork to establish such an active learning culture. It should be deliberately and concretely communicated with specific ideals.
Further prerequisites include a company-wide AI strategy, technical framework conditions, and appropriate resources. To ensure that AI-supported applications function or develop their full effectiveness, a company-wide AI strategy is needed. This strategy should align with the company's vision and the goals of the learning strategy, while also keeping an eye on data flows and system interoperability. Additionally, companies need a robust technical infrastructure that can support AI-based tools and platforms. They must also ensure the privacy and security of learners are guaranteed. Finally, sufficient financial resources and means must be made available to implement and maintain AI systems.
2. Level: Responsible Managers & Leaders
As with any new technology, the success of learning with artificial intelligence depends on technological openness. New technologies can only be understood if one engages with them and tries them out.
Responsible managers and leaders must understand that dealing with AI requires its own learning process. Competence expansion requires providing learners and users with enough time and suitable opportunities.
The transition to learning with AI is facilitated if managers and leaders act as ambassadors and mediators by exemplifying AI learning. This breaks down barriers, builds trust, and democratizes learning with AI.
The responsible individuals must also ensure that the use of AI complies with local and international laws and standards. This includes, for example, compliance with the General Data Protection Regulation (GDPR) and other relevant legal regulations. This can be achieved partly through configuration. Since there may be changes to the laws, continuous monitoring should be conducted. Regarding the processing of personal data, GDPR compliance applies to all AI tools and systems. The principles of data minimization, purpose limitation, and transparency must be adhered to. It must be assessed whether personal data needs to be used at all and, if so, to what extent. Alternatively, data can be pseudonymized or anonymized.
When using generative AI, ethical considerations about fairness, transparency, and avoiding bias become increasingly important. The principle of equal treatment must be maintained, and AI-based decisions must be free of discrimination. When generating texts, images, audios, and videos by AI, it must be ensured that no prejudices are perpetuated and amplified through the dissemination of these contents, but rather that representations are diverse. Companies should develop their own ethical guidelines for using AI to make the work of AI systems fair and transparent and adhere to these. This can help avoid discrimination, exclusion, and the spread of prejudices.
To meet all these points and use the tools effectively, L&D staff must be appropriately qualified. This includes AI literacy, an understanding of opportunities and challenges, and the ability to try out the tools themselves. Only in this way can use cases be identified and implemented in compliance with the law. Ongoing training and exchange with other departments ensure sustainable use. This benefits both the learners and the L&D departments themselves.
3. Level: Users & Learners
Learners need to understand the philosophy of lifelong learning. Learning with AI requires a willingness, openness, and adaptability to engage with new things. This may require stepping out of one's comfort zone. It is also relevant that the use of technologies changes the half-life of specialized knowledge: the more advanced the technology, the shorter the half-life. Currently, this period is three to five years. Users must therefore purposefully repeat and refresh content and be prepared to integrate learning into their daily routines.
Another important factor in learning with AI is feedback for learners. They should consciously choose AI systems that, for example, can provide feedback on the status and development of learning progress.
Learners should be aware that AI can provide individual support and adapt to personalities and special learning needs. This requires trust in the organization, the use of AI, and the learning process. If this trust is present, individual learning goals can be defined, self-directed learning can be enabled, and optimal individual learning support can be offered.
The application areas of AI in companies' L&D departments are diverse. They range from creating learning content to automatically adapting courses to the learner's learning level, to analyzing learning progress data. The following are various use cases:
In personalized learning, AI provides a tailored learning opportunity that adapts to the user's role and offers individual recommendations.
Adaptive learning goes even further, ranging from simple pre-tests to advanced machine learning algorithms. Using these data and algorithms, courses and content are tailored to the learners' needs. In particularly advanced systems, users can learn as tailored as with a personal tutor. Individual adjustments can be made before, during, and after a course.
Another scenario is support in employee career planning. AI can create personalized learning paths or offer learning recommendations that optimally prepare an employee for a new position. Through benchmarking and development recommendations, employees can be specifically prepared for soft skills and future techniques.
Skill competency management systems can depict and display skills and competencies in employee profiles and learning content. This supports tracking and developing skills and makes them directly visible.
AI can tailor daily learning recommendations to various users. These are based on the individual learning behavior and preferences of the user. To provide appropriate content, past interactions and learning successes are also analyzed.
LLMs for daily learning can be used as personal learning companions and sparring partners. They can provide users with individual and appropriate examples and explain complex matters in more detail. They can also pose tasks from everyday work to reinforce what has been learned.
Learners can receive real-time feedback on their learning progress through AI. This includes completed courses, achieved learning objectives, and other factors. By analyzing data and recognizing patterns, the strengths and weaknesses of the user can be identified. Based on this, feedback and suggestions for improvement can be provided.
Generative AI can be used to support the L&D department. For example, employees can create a structure for an e-learning course from designed learning content. In addition, AI can generate scripts for videos, images, or entire videos, thus supporting the production of e-learning. For already created content, AI can create quiz questions to test learning.
How generative AI can further support employees in L&D departments is summarized in the following cheat sheet. The possibilities are constantly growing, and new tools and applications are added daily.
You can find a detailed explanation in the cheat sheet of our guideline.
