(de-news.net) – An AI tax at an early stage has been rejected by Germany’s Digital Minister Karsten Wildberger (CDU), who prioritizes economic growth and domestic innovation. Meanwhile, an Ifo Institute study warns that restrictions on AI training data could reduce Germany’s potential annual economic gains by €155 billion by 2034. The findings highlight the importance of data access, regulatory certainty and domestic technological development, as policymakers weigh AI’s economic opportunities. Separately, German employers’ chief Rainer Dulger has urged swift government reforms to reduce bureaucracy and business costs, strengthen competitiveness and advance the proposed pension reform.

Minister Wildberger has ruled out that the Federal Government might introduce an immediate tax on artificial intelligence, arguing that the country should prioritize economic growth and domestic innovation before considering additional levies. In an interview with German broadcasters RTL and ntv, Wildberger said the government should first allow AI-driven growth to generate additional revenue rather than introduce a new tax at this stage. The government would continue to examine the appropriate regulatory framework and consider potential fiscal measures as the technology’s economic impact develops.

Wildberger stressed that Germany must first strengthen its domestic capacity to generate digital economic value through technological development, platform businesses and payment systems. The country has failed to capture a sufficient share of the value created in these sectors, he argued, resulting in lost employment opportunities and tax revenue. Building stronger domestic innovation capabilities should therefore be a central priority for future economic growth. Wildberger had previously told the German newspaper Der Tagesspiegel that the substantial revenues generated by AI could eventually be taxed to help finance the restructuring of the labor market. Asked whether the burden should fall primarily on major technology companies, he indicated that the proposal remained open for discussion.

Wildberger also expects AI to bring fundamental changes to the labor market, with employment patterns shifting as the technology develops. Although new jobs may eventually offset some of the positions affected by automation and other technological changes, he said those opportunities would emerge in different occupations from many of those that exist today. Retraining and transitions into professions such as skilled trades and nursing will consequently require financial support. According to Wildberger, economic growth will be essential to meeting those costs, although other policy instruments could also be considered.

Data access seen as key to Germany’s AI growth potential

Meanwhile, a study by the Ifo Institute has cautioned that restrictions on access to AI training data could substantially reduce Germany’s potential economic gains from the technology. Generative AI could generate an additional 440 billion euros in annual domestic value added by 2034, according to the institute’s model calculation. That potential increase would be reduced by 155 billion euros if AI developers faced greater difficulties obtaining the data needed to train their models. The findings highlight the economic significance of data access as policymakers consider how copyright rules and other regulatory requirements should apply to AI development.

The projected economic benefits consist of approximately 330 billion euros from the widespread adoption of AI, 65 billion euros from productivity gains in research, and 45 billion euros from AI development within Germany. Oliver Falck, head of the Ifo Center for Innovation Economics and Digital Transformation, and his co-author Martin Thelle of Implement Consulting Group identified four principal channels through which tighter restrictions could diminish those gains. Slower adoption could cost 45 billion euros, while the development of less capable models could account for a further 75 billion euros in lost value. Reduced research potential was estimated at 25 billion euros, and weaker AI development in Europe at 10 billion euros.

The researchers emphasized that advanced AI language models depend on access to enormous and diverse quantities of training data. Europe’s largest copyright-free dataset contains only two trillion to three trillion tokens, compared with the 15 trillion to 36 trillion tokens already used by leading international models. The disparity illustrates the scale of the data challenge facing European developers as they seek to compete in the global AI market. To reduce uncertainty for investors and support further technological development, the authors advocated retaining the text-and-data-mining exception established under the European Union’s Copyright Directive.

Separately, Rainer Dulger, president of the German Employers’ Association, called on the Federal Government to make Germany’s economy more competitive, rejecting suggestions that businesses themselves had failed to undertake necessary reforms. Speaking to RTL and ntv, Dulger argued that the country’s economic difficulties stemmed primarily from outdated framework conditions rather than a failure by companies to recognize the need for change. He identified excessive bureaucracy, payroll-related costs and government structures as areas requiring reform, emphasizing that Germany needed to become faster, more efficient and simpler for businesses. Employers, he added, were prepared to support the government’s reform efforts wherever possible.