Simon Morris’s core message was that European enterprises should not define AI sovereignty narrowly as eliminating American technology, but as deciding where they need control, where dependence is acceptable, and how they protect the value AI will create. He opened by saying he would “love to see a German AWS or a French Azure or even a British OpenAI,” but argued that removing all American tech from European businesses is not realistic in the short term. Instead, as AI becomes more integral to the industrial economy, leaders should focus on the specific points where their organizations can retain control and where reliance on external parties does not undermine strategic capacity.

He grounded the argument in Germany’s industrial history, describing the country as defined for more than a century by its ability to master machines, create cars, automotive parts, and industrial excellence. Morris argued that as AI enters manufacturing and industrial processes, differentiation shifts: the unique value is no longer just raw inputs, process excellence, or finished components, but the intelligence generated during production. For him, AI sovereignty means capturing and using that intelligence rather than letting it leak to companies over which European firms have no control or leverage.

Morris then expanded sovereignty beyond nationality. He said the “flag over the headquarters” is visible but too narrow, and that the real issue is dependencies: supply-chain choke points, IP control, critical production capacity, export controls, tariffs, and the ability to turn capacity on or off. He proposed a three-part framework for business leaders: control, substitution, and recovery. Control asks what parts of the business require authority; substitution asks whether critical capabilities can be moved to another supplier; recovery asks how quickly access to capacity or components can be restored after disruption.

Applying this to AI, Morris used Jensen Huang’s “five layer cake” model: energy, chips, data centers, models, and business value. He emphasized that energy and chips are deeply interdependent, and then examined microchips as a sovereignty challenge. TSMC’s ESMC initiative in Europe and Intel’s abandoned German manufacturing plan illustrated that fabrication alone does not create sovereignty if Europe lacks IP ownership, capacity control, and a customer ecosystem. In the Q&A, an audience member challenged Germany’s basic infrastructure and energy readiness; Morris responded that human adoption is today’s AI choke point, but power will become a fundamental constraint because without energy, chips, data centers, models, and future value cannot scale.