What are the best-to-use certifications

Certification of Artificial Intelligence: Platform for Learning Systems identifies challenges

Artificial intelligence (AI) is already used in many industries. In order to exploit their economic and social potential, it is essential to strengthen trust in AI systems and the processes and decisions associated with them. A possible key requirement for this is certification of AI systems. In an impulse paper, experts from the Platform for Learning Systems outline what benefits it promises and what requirements arise with a view to technical implementation, the common good and the preservation of innovative strength. It gives an overview of existing certification projects in Germany and forms the basis for further discussions.

Particularly in sensitive areas of application such as medicine, certification of AI systems can help to increase confidence in their performance, reliability and security. In a business context, certification facilitates the interoperability of different systems and thus promotes the continued use of artificial intelligence. Last but not least, certification could promote competitive dynamics in the development of AI applications and - by establishing a trustworthy brand "AI made in Europe" - create international competitive advantages.

Special challenges in learner systems

However, there are numerous questions to be clarified on the way to certification of Artificial Intelligence. "For example, it must be answered how learner systems can be reliably verified, or how further learning can be ensured in the company - for example through structured updates", says Prof. Dr. Stefan Wrobel, Head of the Fraunhofer Institute for Intelligent Analysis and Information Systems IAIS and member of the Technological Trailblazers and Data Science working group of the Learning Systems platform. “Another challenge is that AI applications are often hybrid systems - that is, they are based on a combination of different AI technologies. In the field of language technology, for example, machine learning processes are often combined with model knowledge. Such complex systems must also be covered by certification, ”says Stefan Wrobel.

Finding the right level is also crucial for a meaningful and beneficial certification of AI systems. “The task is to develop a general test system in order to make a certification of highly different AI systems for different areas of application comparable. It is important to consider already established norms and standards and to close existing gaps, ”explains Stefan Wrobel, one of four co-authors of the impulse paper.

Finding the right level for certification

The impulse paper “Certification of AI systems” created by the learning systems platform by an interdisciplinary team of authors illuminates not only technical, but also legal and ethical aspects. “For a large number of AI systems, certification can help to exploit their potential societal benefits safely and in a way that is oriented towards the common good. In order for this to happen in harmony with socially recognized values, a form of certification must be found that is guided by important ethical principles, but at the same time also meets economic principles, avoids overregulation and promotes innovation, ”says Jessica Heesen, head of the research focus on media ethics and information technology at the International Center for Ethics in Science (IZEW) at the University of Tübingen and head of the IT security, privacy, law and ethics working group of the learning systems platform. The co-author of the impulse paper adds: "In the best case scenario, certification itself can trigger new developments for a European path in AI application."

About the impulse paper

The impulse paper Certification of AI systems illuminates the potential and challenges that the certification of AI systems entails and gives an overview of existing certification projects in Germany. The paper was written by members of the platform learning systems under the leadership of the IT security, privacy, law and ethics working group and the technological trailblazers and data science working group and is intended as a basis for further discussions on the topic.

Back