TechBBQ puts European AI sovereignty at the centre of debate
Control of AI became the central question at TechBBQ
At TechBBQ in Copenhagen, investors, founders and operators from across Europe repeatedly returned to one issue: who controls the technology powering artificial intelligence. The annual Nordic conference adopted the theme “Emerging from Agency”, and discussions ranged beyond what organisations can build with AI to the degree of control they retain over models, data and infrastructure.
The question gained urgency after Anthropic’s AI models Mythos and Fable became unavailable to users outside Europe earlier this year. The disruption prompted participants to consider the consequences of relying on AI capabilities supplied by companies and infrastructure based in the United States and China. One startup executive said the incident seriously disrupted a software team, while another attendee took a less immediate view of the risk.
Agency, privacy and concentrated power
Ellen de Brever, angel investor and head of partnerships at the Novo Nordisk Foundation Cellerator, said the conference theme was timely because the focus had shifted from AI’s technical capabilities to the choices people make about its use. She said the debates were not about smarter machines, but about human judgement and “who gets to stay in the driver’s seat”.
De Brever characterised the age of AI agents as a question of humans deciding what to relinquish, rather than machines independently gaining agency. This framing connected discussions on AI to panels on women’s health, Nordic and African innovation and investment, and Europe’s need for more growth capital.
Signal President Meredith Whittaker addressed the relationship between privacy and AI in a session on whether the two can coexist. She criticised the integration of AI assistants and agents into operating systems, including ChatGPT’s work with iMessage, and described the emerging environment as a “data collection apparatus”. Whittaker argued that AI labs’ marketing can obscure the consequences of collecting large quantities of data, while maintaining that demand for privacy remains substantial.
On another panel, Stability AI and Intelligent Internet co-founder Emad Mostaque discussed an agentic workforce, personal sovereignty and the concentration of power among a small number of AI labs. He defined sovereignty as the ability to resist power exerted over an individual, and argued that control of AI could become consequential at country level.
What businesses should take from the discussion
TechBBQ’s conversations did not produce a single definition of European AI sovereignty, but they made dependencies visible. The availability of Anthropic’s Mythos and Fable models showed how access to externally controlled AI services can affect software teams, while privacy concerns highlighted the significance of where data is collected and how assistants are integrated.
For businesses adopting AI, the practical implication is to identify which provider controls each critical model, data flow and infrastructure component, and to understand the operational impact if access changes. Those questions belong alongside decisions about capability, cost and deployment.