Dr. Karsten Wildberger’s core message was that AI can become Germany’s leapfrog moment if companies and government treat it as a strategic capability rather than a set of isolated applications. He opened by contrasting two Mittelstand experiences: one company adopted AI but still failed to remove its IT development bottleneck, while another used AI agents to replace a legacy CRM-related core system within ten weeks. From that contrast, he argued that every company has a different starting point, but the practical question should be how AI could reshape the business model itself, including how a new AI-native competitor would attack an incumbent’s position.

He then reframed the AI risk debate around three categories. The first two were familiar: misuse by criminals or rogue actors, including cyberattacks or biological weapons, and loss of control, where autonomous agents may pursue unintended goals. Wildberger stressed that companies cannot treat the world as a test bed, should test before shipping, accept independent scrutiny, and remain liable after deployment. The third risk, which he said receives too little attention in Germany, was dependency and lost value creation. If Germany does not build its own AI capabilities and competitive products, he warned that businesses, skilled jobs, value creation, growth, wealth, and independence will move elsewhere.

To show that the government is acting, Wildberger cited AC Deutschland, Germany’s AI safety institute being established with the Interior Ministry; support for Law Zero, Joshua Bengio’s nonprofit developing safe-by-design foundation models; and a data center strategy intended by 2030 to at least double overall compute capacity and quadruple AI compute capacity. He also pointed to sovereign cloud infrastructure for the federal system, including German consortia involving SAP with Telecom and Schwarz Digits, as a backbone for digitization and protected environments where Mittelstand companies can use AI while retaining control over industrial knowledge.

The speech moved from risk to opportunity by emphasizing industrial AI, frontier model partnerships, and government procurement as demand creation. Wildberger highlighted the combination of Cohere and Aleph Alpha as potentially one of the largest large language model companies outside the US and China, noted Mistral’s Munich office and partnership with SAP, and described SPRK, government AI agents that automate complex infrastructure approvals and cut time and effort by about 85%. In the Q&A opening, the moderator cited a BITKOM study saying 78% of Mittelstand and large corporations in Germany see AI as decisive for competitiveness, while nearly 50% say they are lagging behind; Wildberger responded that adoption must happen much faster, noting that AI diffusion appears slower globally outside China.