At Copenhagen’s TechBBQ conference, European investors, founders and operators repeatedly debated who should control the AI systems shaping business and society. The discussion centred on technological sovereignty: Europe’s dependence on US and Chinese models and infrastructure, and the risk that external providers could restrict access or influence how AI is deployed. Speakers also framed the issue as one of human agency, asking what decisions people are willing to delegate to AI systems.

The concern was sharpened by Anthropic’s Mythos and Fable models becoming unavailable outside Europe earlier this year, which disrupted at least one startup team. Signal president Meredith Whittaker warned that AI assistants integrated into operating systems could create extensive data-collection systems, while Stability AI co-founder Emad Mostaque argued that concentrated AI control could translate into national power. Other attendees said widespread AI agents and robotics may eventually require societies to reconsider ownership, work, democracy and participation.