What rules does AI need? Plattform Lernende Systeme provides an overview
Munich, 15 May 2023
Artificial Intelligence (AI) is a key technology for a sustainable economy and society. At the same time, AI systems also harbor risks. In the current debate about ChatGPT and other generative AI models, there are calls for regulation of the technology. An overview of possible forms and current regulatory initiatives – from the European AI Act to industry standards – is provided by a new web service of Plattform Lernende Systeme.
Whether government standards such as the currently negotiated AI Act of the European Union (EU) or voluntary commitments by industries or individual companies – the goal of successful rules for Artificial Intelligence is to design safe, trustworthy AI systems. At the same time, uniform rules create fair competitive conditions for business.
How can we design safe and trustworthy AI systems? What rules for AI development and deployment do we need to minimize risks? And what is the right level of regulation to support competitive AI innovations from Europe? The new web service from Plattform Lernende Systeme provides impetus to these questions – with viewpoints from its experts on AI regulation, literature tips for further reading, and an overview of existing guides and test catalogs for regulating AI systems. The EU’s AI Act and the German Institute for Standardization’s (DIN) AI Standardization Roadmap are detailed as key regulatory initiatives.
“Standards and norms for AI systems protect people as well as companies from potential risks associated with the use of AI applications. However, it is important to strike the right balance between imposing tough restrictions and promoting innovation for Germany as a business location. Overregulation must be avoided at all costs The goal is to shape the technology in a way that inspires confidence. In the best case, we will succeed in creating a kind of quality seal ‘AI made in Europe’ through uniform rules in Europe,” says Jan Wörner, co-chair of Plattform Lernende Systeme.
The website can be accessed here (in German). All statements by the experts have been released for editorial use.