Germany Says Halting AI Development Not ‘Viable’ as EU Weighs Safety and Competitiveness

Germany on Monday pushed back against mounting calls to temporarily halt the development of advanced artificial intelligence systems, stating that such a pause was not a “viable option” for Europe. The statement comes as governments across the continent grapple with how to regulate powerful AI models without sacrificing competitiveness in a global race increasingly defined by the United States and China.

“Halting the development of artificial intelligence systems is not a viable option for Europe,” German officials said in a statement. Instead, they argued, the focus must shift toward managing the risks of the technology while fostering innovation.

Europe’s AI dilemma: safety versus sovereignty

The German position underscores a deepening tension within the European Union, where the pioneering EU AI Act — the world’s first comprehensive regulatory framework for artificial intelligence — is already being implemented. While the Act imposes strict transparency and safety requirements on high-risk AI applications, some member states and industry players fear that overly rigid rules could stifle innovation and leave Europe dependent on foreign technology.

Germany’s latest intervention signals a clear preference for a governance model that does not throttle AI development outright but rather steers it through a combination of safety protocols, international cooperation, and industrial policy. Berlin is reportedly urging that the conversation broaden to include not just EU member states but also global powers like the U.S. and China, whose AI ecosystems dwarf those of most European nations.

The limits of a moratorium

Calls for a temporary halt to AI development have gained traction in recent years, most famously in an open letter signed by tech leaders and researchers in early 2023 that demanded a six-month pause on training models more powerful than GPT-4. Civil society groups and prominent AI ethicists have also warned that unfettered progress could lead to runaway risks. However, that proposal has since been met with skepticism from many policymakers, who argue that a moratorium is impossible to enforce and would merely hand the advantage to less regulated competitors.

Germany appears to be aligning itself with that realist perspective. Rather than a freeze, Berlin is pushing for what it describes as “viable” alternatives: compulsory risk assessments, independent auditing of frontier AI systems, and international agreements that set baseline safety standards. The government has also stressed the need to keep AI development in Europe dynamic enough to attract and retain talent, investment, and critical infrastructure.

Instead of a moratorium, German officials have pointed to concrete measures such as mandatory red-teaming, third-party safety evaluations, and strict liability rules for AI-generated harm. Berlin has also championed the concept of “regulatory sandboxes” that allow companies to test new AI products in controlled environments before full-scale deployment — an idea already embedded in the AI Act.

Broadening the governance table

A key element of Germany’s approach is the involvement of both the United States and China in global AI governance discussions. While the U.S. has been a close ally in many tech policy debates, China represents a far more complex challenge given its centralized AI ambitions and differing views on censorship and surveillance. Nevertheless, German officials have argued that leaving Beijing out of global safety frameworks would make those frameworks inherently incomplete.

This push for inclusivity is reflected in ongoing work at international bodies. The OECD AI Policy Observatory already tracks over 1,000 AI policy initiatives across 70 countries, and the UNESCO Recommendation on the Ethics of Artificial Intelligence, adopted in 2021, is the first global normative instrument in the field. Germany’s stance aligns with these multilateral efforts, advocating for a patchwork of safeguards rather than a one-size-fits-all ban.

Industry and geopolitical stakes

The AI sector has become a central pillar of geopolitical competition. The United States leads in foundation model development through companies like OpenAI, Google DeepMind, and Anthropic, while China has made rapid advances with Baidu’s Ernie Bot and state-backed initiatives. Europe, by contrast, has a robust regulatory posture but lags in scaling up homegrown AI champions capable of rivaling American or Chinese counterparts.

Germany’s Economy Ministry has previously outlined plans to invest billions in AI research and chip manufacturing, hoping to bolster the continent’s “digital sovereignty.” A moratorium, in Berlin’s view, would directly undermine those ambitions by freezing progress just as European firms begin to catch up.

Observers note that the German stance may influence the EU’s broader direction as it prepares to review and possibly recalibrate the AI Act in the years ahead. While the legislation already prohibits certain AI practices deemed unacceptable — such as social scoring and untargeted facial recognition scraping — it stops short of banning entire models or halting research. The latest German statements reinforce that philosophy, insisting that safety and innovation are not mutually exclusive.

What comes next

The discussion is set to continue in multiple forums this autumn, including a high-level EU summit on AI and competitiveness and a G7 meeting where digital ministers will address cross-border AI risks. Germany is expected to use these platforms to advance its call for a globally coordinated risk-management framework, rather than a single regulatory chokehold.

In the meantime, AI development races ahead. Whether Europe can strike the balance between caution and competitiveness remains an open question — but for Germany, pausing progress altogether is not the answer.