Dispute over opportunities and risks of AI at digital summit
Different positions on the regulation of artificial intelligence (AI) clashed at the German government's digital summit. While Federal Ministers Robert Habeck (Greens) and Volker Wissing (FDP) were firmly against restricting the basic technology, representatives of civil society called for strict legal framework conditions not only for specific AI applications, but also for the basic technology.
Matthias Spielkamp, co-founder and managing director of the non-governmental organization AlgorithmWatch, referred to a recent study of the responses of AI software robots to the recent state elections in Hesse and Bavaria. "The systems spit out a lot of nonsense." In this example, not only had the companies developed models that could have very negative effects, but the companies had already brought them onto the market. The self-regulation proposed by the German government for basic technology would not do justice to the challenges posed by AI.
Habeck: "Everything is susceptible to abuse"
Habeck, on the other hand, defended the German government's position of making a distinction between the basic technology and the specific application of AI. "Everything is susceptible to abuse. We can use electricity to kill people and build electric chairs, and we can use electricity to operate healthcare devices or make access to medicine and education easier and better."
Before you can commit a technology to upholding social values, you first have to have that technology, said Habeck. "We have to be careful that regulation does not restrict us so much that in the end only Elon Musk remains." The American tech billionaire founded his AI company xAI in July and recently presented his first AI program "Grok".
Carla Hustedt, Head of Digitalized Society at the Mercator Foundation, warned against too lax regulation: "We must not follow the Chinese or American path just because we are afraid of being left behind."
During the debate, Digital and Transport Minister Wissing pointed out that it is not yet possible to predict where AI will be next year or the year after. "We don't have to finalize all issues today. We can make adjustments at any time."
Negotiations on AI are still ongoing in the EU
In the European Union, decisive negotiations on the new AI Act are currently underway between the European Council, the European Parliament and the European Commission. An agreement should be reached by the end of the year. Germany recently agreed on a joint position paper with France and Italy, which advocates regulation for AI applications. When it comes to basic technologies, the three largest EU states want to limit themselves to self-regulation of the industry.
A number of non-governmental organizations and representatives of civil society were also invited to the digital summit in Jena for the first time. Against this backdrop, Christian Humborg, Managing Director of Wikimedia Deutschland, spoke of a "positive development towards more participation and diversity". The digital summit was better than last year: at the same time, Humborg criticized a "fragmentation of responsibilities and competencies" in Germany. This means that there is no discernible signature in terms of content, nor does digitization work in practice. "You can't have a complete digital policy with half a digital ministry."
Markus Beckedahl, founder of the online platform Netzpolitik.org, spoke of a "fragile start to the second half of the German government's legislative term, which has so far only largely disappointed on digital policy issues". "But it should also be clear: We don't just need a symbolic digital summit once a year, we need to continuously discuss and negotiate the shaping of the digital world - and not just behind closed doors with industry representatives, but with society as a whole," said Beckedahl.
The Federal Government needs to consider the role of information technology and telecommunications in regulating AI, as recent instances have shown that unrestricted AI technology can lead to the release of nonsensical information on platforms like the Internet.
During the negotiations on the EU's new AI Act, the German government, along with France and Italy, has advocated for self-regulation for basic AI technologies by the industry, while pushing for regulation of AI applications.
Source: www.dpa.com
